PARAGRAPH PAGE

A.1 GENERAL. 50A.1.1 Scope. 50A.1.2 Applicability. 50A.2 APPLICABLE DOCUMENTS. 50A.2.1 General. 50A.2.2 Government documents. 50A.2.2.1 Specifications, standards, and handbooks. 50A.2.3 Non-Government publications. 51A.3 DEFINITIONS. 52A.3.1 Terms. 52A.3.2 Abbreviations and acronyms. 52A.3.3 Definitions of timing symbols. 54A.4 GENERAL REQUIREMENTS. 55A.4.1 ALE introduction. 55A.4.1.1 ALE addresses. 55A.4.1.2 Scanning. 55A.4.1.3 Calling. 56A.4.1.4 Channel evaluation. 56A.4.1.5 Channel quality display. 57A.4.2 System performance requirements. 57A.4.2.1 Scanning rate. 57A.4.2.1.1 Alternative Quick Call (AQC). 57A.4.2.1.2 Recommendation. 57A.4.2.2 Occupancy detection - not tested (NT). 57A.4.2.3 Linking probability. 58A.4.2.3.1 AQC-ALE linking probability. 60A.4.2.3.2 AQC-ALE linking performance. 60A.4.3 Required data structures. 60A.4.3.1 Channel memory. 60A.4.3.2 Self address memory. 61A.4.3.3 Other station table. 63A.4.3.3.1 Other station address storage. 63A.4.3.3.2 Link quality memory. 65A.4.3.3.3 Other station settings storage. 65A.4.3.4 Operating parameters. 65A.4.3.5 Message memory. 65A.4.4 ALE operational rules. 66A.4.5 Alternate Quick Call ALE (AQC-ALE) (NT). 66A.4.5.1 Introduction. 66A.4.5.2 General signaling strategies. 66TABLE OF CONTENTS(continued)PARAGRAPH PAGEA.4.5.3 Features supported by AQC-ALE. 67A.4.5.4 Features not provided by AQC-ALE. 68A.5. DETAILED REQUIREMENTS. 68A.5.1 ALE modem waveform. 68A.5.1.1 Introduction. 68A.5.1.2 Tones. 68A.5.1.3 Timing. 69A.5.1.4 Accuracy. 69A.5.2 Signal structure. 71A.5.2.1 Introduction. 71A.5.2.2 FEC. 71A.5.2.2.1 General. 71A.5.2.2.2 Golay coding. 71A.5.2.2.2.1 Encoding. 71A.5.2.2.2.2 Decoding. 71A.5.2.2.3 Interleaving and deinterleaving. 76A.5.2.2.4 Redundant words. 76A.5.2.3 Word structures. 76A.5.2.3.1 ALE word format. 76A.5.2.3.1.1 Structure. 81A.5.2.3.1.2 Word types. 81A.5.2.3.1.3 Preambles. 81A.5.2.3.2 Address words. 82A.5.2.3.2.1 TO. 82A.5.2.3.2.2 THIS IS (TIS). 82A.5.2.3.2.3 THIS WAS (TWAS). 82A.5.2.3.2.4 THRU. 82A.5.2.3.2.5 FROM. 83A.5.2.3.3 Message words. 83A.5.2.3.3.1 CMD. 83A.5.2.3.4 Extension words. 84A.5.2.3.4.1 DATA. 84A.5.2.3.4.2 REP. 84A.5.2.4 Addressing. 84A.5.2.4.1 Introduction. 84A.5.2.4.2 Basic 38 subset. 85A.5.2.4.3 Stuffing. 86A.5.2.4.4 Individual addresses. 86A.5.2.4.4.1 Basic size. 88A.5.2.4.4.2 Extended size. 88TABLE OF CONTENTS(continued)PARAGRAPH PAGEA.5.2.4.5 Net addresses. 90A.5.2.4.6 Group addresses. 90A.5.2.4.7 Allcall addresses. 90A.5.2.4.8 AnyCalls. 91A.5.2.4.9 Wildcards. 91A.5.2.4.10 Self addresses. 92A.5.2.4.11 Null address. 93A.5.2.4.12 In-link address. 93A.5.2.5 Frame structure. 93A.5.2.5.1 Calling cycle. 95A.5.2.5.2 Message section. 98A.5.2.5.3 Conclusion. 100A.5.2.5.4 Valid sequences. 103A.5.2.5.5 Basic frame structure examples. 103A.5.2.6 Synchronization. 107A.5.2.6.1 Transmit word phase. 107A.5.2.6.2 Receiver word sync. 108A.5.2.6.3 Synchronization criteria. 108A.5.3 Sounding. 109A.5.3.1 Introduction. 109A.5.3.2 Single channel. 110A.5.3.3 Multiple channels. 110A.5.3.4 Optional handshake. 115A.5.4 Channel selection. 117A.5.4.1 LQA. 117A.5.4.1.1 BER. 117A.5.4.1.2 SINAD. 118A.5.4.1.3 MP (optional). 118A.5.4.1.4 Operator display (optional). 118A.5.4.2 Current channel quality report (LQA CMD). 118A.5.4.2.1 BER field in LQA CMD. 118A.5.4.2.2 SINAD. 118A.5.4.2.3 MP. 118A.5.4.3 Historical LQA report. 120A.5.4.4 Local noise report CMD (optional). 120A.5.4.5 Single-station channel selection. 121A.5.4.5.1 Single-station channel selection for link establishment. 121A.5.4.5.2 Single-station channel selection for one-way broadcast. 123A.5.4.5.3 Single-station channel selection for listening. 123A.5.4.6 Multiple-station channel selection. 123TABLE OF CONTENTS(continued)PARAGRAPH PAGEA.5.4.7 Listen before transmit. 124A.5.4.7.1 Listen-before-transmit duration. 124A.5.4.7.2 Modulations to be detected. 124A.5.4.7.3 Listen before transmit override. 124A.5.5 Link establishment protocols. 124A.5.5.1 Manual operation. 124A.5.5.2 ALE. 125A.5.5.2.1 Timing. 125A.5.5.2.2 ALE states. 125A.5.5.2.3 ALE channel selection. 125A.5.5.2.3.1 Rejected channel. 126A.5.5.2.3.2 Busy channel. 126A.5.5.2.3.3 Exhausted channel list. 126A.5.5.2.4 End of frame detection. 126A.5.5.3 One-to-one calling. 129A.5.5.3.1 Sending an individual call. 129A.5.5.3.2 Receiving an individual call. 130A.5.5.3.3 Response. 131A.5.5.3.4 Acknowledgment. 132A.5.5.3.5 Link termination. 133A.5.5.3.5.1 Manual termination. 133A.5.5.3.5.2 Automatic termination. 133A.5.5.3.6 Collision detection. 134A.5.5.4 One-to-many calling. 134A.5.5.4.1 Slotted responses. 134A.5.5.4.1.1 Slotted response frames. 135A.5.5.4.1.2 Slot widths. 135A.5.5.4.1.3 Slot wait time formula. 135A.5.5.4.1.4 Slotted response example. 136A.5.5.4.2 Star net calling protocol. 136A.5.5.4.2.1 Star net call. 136A.5.5.4.2.2 Star net response. 137A.5.5.4.2.3 Star net acknowledgment. 137A.5.5.4.3 Star group calling protocol. 137A.5.5.4.3.1 Star group scanning call. 137A.5.5.4.3.2 Star group leading call. 138A.5.5.4.3.3 Star group call conclusion. 138A.5.5.4.3.4 Receiving a star group call. 138A.5.5.4.3.5 Star group slotted responses. 139A.5.5.4.3.6 Star group acknowledgment. 139TABLE OF CONTENTS(continued)PARAGRAPH PAGEA.5.5.4.3.7 Star group call example. 139A.5.5.4.3.8 Multiple self addresses in group call. 139A.5.5.4.4 Allcall protocol. 140A.5.5.4.5 AnyCall protocol. 141A.5.5.4.6 Wildcard calling protocol. 142A.5.6. ALE control functions (CMDs other than AMD, DTM, and DBM). 142A.5.6.1 CRC. 144A.5.6.2 Power control (optional). 146A.5.6.3 Channel related functions. 147A.5.6.3.1 Channel designation. 147A.5.6.3.2 Frequency designation. 148A.5.6.3.3 Full-duplex independent link establishment (optional). 149A.5.6.3.4 LQA polling (optional). 149A.5.6.3.5 LQA reporting (optional). 149A.5.6.3.6 LQA scan with linking (optional). 149A.5.6.3.7 Advanced LQA (optional). 149A.5.6.4 Time-related functions. 149A.5.6.4.1 Tune and wait. 149A.5.6.4.2 Scheduling commands. 150A.5.6.4.3 Time exchange word formats. 154A.5.6.4.3.1 Command words. 154A.5.6.4.3.2 Time Is command. 154A.5.6.4.3.3 Time Request command. 154A.5.6.4.3.4 Other encodings. 154A.5.6.4.4 Coarse time word. 154A.5.6.4.5 Authentication word. 155A.5.6.4.6 Time quality. 155A.5.6.5 Mode control functions (optional). 157A.5.6.5.1 Modem negotiation and handoff. 158A.5.6.5.1.1 Modem selection CMD. 158A.5.6.5.1.2 Modem negotiating. 158A.5.6.5.2 Crypto negotiation and handoff. 159A.5.6.6 Capabilities reporting functions. 160A.5.6.6.1 Version CMD (mandatory). 160A.5.6.6.2 Capabilities function. (mandatory). 161A.5.6.6.2.1 Capabilities query. 161A.5.6.6.2.2 Capabilities report CMD. 162A.5.6.6.2.3 Data format. 162A.5.6.7 Do not respond CMD. 165A.5.6.8 Position report (optional). 165TABLE OF CONTENTS(continued)PARAGRAPH PAGEA.5.6.9 User unique functions (UUFs). 165A.5.7 ALE message protocols. 167A.5.7.1 Overview. 167A.5.7.2 AMD mode (mandatory). 167A.5.7.2.1 Expanded 64-channel subset. 167A.5.7.2.2 AMD protocol. 168A.5.7.2.3 Maximum AMD message size. 169A.5.7.3 DTM mode. 169A.5.7.4 DBM mode. 181A.5.8 AQC (optional) (NT). 194A.5.8.1 Signaling structure (NT). 194A.5.8.1.1 AQC-ALE word structure (NT). 194A.5.8.1.1.1 Packed address (NT). 194A.5.8.1.1.2 Address differentiation flag (NT). 196A.5.8.1.2 Preambles (NT). 196A.5.8.1.2.1 TO (NT). 196A.5.8.1.2.2 THIS IS (TIS) (NT). 196A.5.8.1.2.3 THIS WAS (TWAS) (NT). 196A.5.8.1.2.4 PART2 (NT). 196A.5.8.1.2.5 INLINK (NT). 197A.5.8.1.2.6 COMMAND (NT). 197A.5.8.1.2.7 DATA (NT). 197A.5.8.1.2.8 REPEAT (NT). 197A.5.8.1.3 AQC-ALE address characteristics (NT). 197A.5.8.1.3.1 Address size (NT). 197A.5.8.1.3.2 Address character set (NT). 197A.5.8.1.3.3 Support of ISDN (option) (NT). 197A.5.8.1.3.4 Over-the-air address format (NT). 197A.5.8.1.4 Address formats by call type (NT). 197A.5.8.1.4.1 Unit addresses (NT). 197A.5.8.1.4.2 StarNet addresses (NT). 198A.5.8.1.4.3 Group addresses (NT). 198A.5.8.1.4.4 AllCall address (NT). 198A.5.8.1.4.5 AnyCall address (NT). 198A.5.8.1.5 Data exchange field (NT). 198A.5.8.1.5.1 DE(1) no data available (NT). 199A.5.8.1.5.2 DE(2) number of to's left in calling cycle (NT). 199A.5.8.1.5.3 DE(3) Inlink resource list (NT). 199A.5.8.1.5.4 DE(4) local noise report (NT). 200A.5.8.1.5.5 DE(5) LQA variation (NT). 201TABLE OF CONTENTS(continued)PARAGRAPH PAGEA.5.8.1.5.6 DE(6) LQA measurement (NT). 202A.5.8.1.5.7 DE(7) number of Tis/Twas left in sounding cycle (NT). 203A.5.8.1.5.8 DE(8) inlink data definition from INLINK (NT). 203A.5.8.1.5.9 DE(9) Inlink data definition from PART2 (NT). 204A.5.8.1.6 PSK tone sequence (optional) (NT). 205A.5.8.1.6.1 PSK tone sequence placement (NT). 205A.5.8.1.6.2 PSK tone sequence generation (NT). 205A.5.8.2 AQC-ALE frame structure and protocols (NT). 205A.5.8.2.1 Calling cycle (NT). 205A.5.8.2.2 Unit call structure (NT). 207A.5.8.2.3 Star net call structure (NT). 207A.5.8.2.4 AllCall frame formats (NT). 208A.5.8.2.5 AnyCall frame formats (NT). 209A.5.8.2.6 Sounding (NT). 210A.5.8.2.7 Inlink transactions (NT). 210A.5.8.2.7.1 Inlink transaction as an acknowledgement (NT). 211A.5.8.2.7.2 CRC for Inlink event sequences (NT). 211A.5.8.2.7.3 Use of address section (NT). 212A.5.8.2.7.4 Slotted responses in an Inlink state (NT). 212A.5.8.3 AQC-ALE orderwire functions (optional) (NT). 212A.5.8.3.1 Operator ACK/NAK transaction command section (optional) (NT). 212A.5.8.3.2 AQC-ALE control message section (optional) (NT). 213A.5.8.3.2.1 AMD dictionary message (NT). 214A.5.8.3.2.2 Channel definition (NT). 217A.5.8.3.2.3 Slot assignment (NT). 218A.5.8.3.2.4 List content of database (NT). 218A.5.8.3.2.5 List database activation time (NT). 218A.5.8.3.2.6 Set database activation time (NT). 218A.5.8.3.2.7 Define database content (NT). 219A.5.8.3.2.8 Database content listing (NT) 220A.5.8.4 AQC-ALE linking protection (NT). 220

TABLE A-I. Occupancy detection probability (2G and 3G). 57TABLE A-II. Probability of linking. 59TABLE A-III. Channel memory example. 62TABLE A-IV. Self address memory example. 63TABLE A-V. ALE operational rules. 66TABLE A-VI. 2/3 majority vote decoding. 79TABLE A-VII. Majority word construction. 79TABLE OF CONTENTS(continued)PARAGRAPH PAGETABLE A-VIII. ALE word types (preambles). 81TABLE A-IX. Use of "@" utility symbol. 87TABLE A-X. Basic (38) address structures. 89TABLE A-XI. Use of "?" wildcard symbol. 92TABLE A-XII. Limits to frames. 103TABLE A-XIII. Basic BER values. 119TABLE A-XIV. Link quality analysis structure. 120TABLE A-XV. Timing. 127TABLE A-XVI. Summary of CMD functions. 143TABLE A-XVII. Cyclic redundancy check structure. 146TABLE A-XVIII. Power control CMD bits (KP1-3). 147TABLE A-XIX. Tune and wait structure. 151TABLE A-XX. Time values. 152TABLE A-XXI. Time-related CMD functions. 153TABLE A-XXII. Time quality. 156TABLE A-XXIII. Modem codes. 159TABLE A-XXIV. Crypto codes. 160TABLE A-XXV. Component selection. 161TABLE A-XXVI. Format selection. 161TABLE A-XXVII. Capabilities report data fields (ALE timing). 163TABLE A-XXVIII. Capabilities report data fields (mode settings). 163TABLE A-XXIX. Capabilities report data field (feature capabilities). 164TABLE A-XXX. User unique functions structure. 166TABLE A-XXXI. ALE message protocols. 167TABLE A-XXXII. DTM characteristics. 171TABLE A-XXXIII. DTM structure. 174TABLE A-XXXIV. DBM characteristics. 182TABLE A-XXXV. DBM structures. 188TABLE A-XXXVI. AQC address character ordinal value. 195TABLE A-XXXVII. AQC-ALE word types (and preambles). 196TABLE A-XXXVIII. Data exchange definitions. 198TABLE A-XXXIX. Inlink resource list. 200TABLE A-XL. Local noise report. 201TABLE A-XLI. Magnitude of minimum SNR from mean SNR. 202TABLE A-XLII. LQA scores. 203TABLE A-XLIII. Valid combinations of ACK-This and I'm Inlink. 204TABLE A-XLIV. DE(9) inlink transaction identifier. 205TABLE A-XLV. Scanning part duration using automated calculation. 206TABLE A-XLVI. Operator ACK/NAK command. 213TABLE A-XLVII. AQC-ALE control message section word sequences. 213TABLE OF CONTENTS(continued)PARAGRAPH PAGETABLE A-XLVIII. Lookup tables for packed AMD messages. 216TABLE A-IL. Adding spaces during AMD unpacking. 216

FIGURE A-1. Data link with ALE and FEC sublayers. 56FIGURE A-2. Occupancy detection test setup. 58FIGURE A-3. System performance measurements test setup. 58FIGURE A-4. Connectivity and LQA memory example. 64FIGURE A-5. ALE symbol library. 70FIGURE A-6. Generator matrix for (24, 12) extended Golay code. 72FIGURE A-7. Parity-check matrix for (24, 12) extended Golay code. 73FIGURE A-8. Golay word encoding example. 74FIGURE A-9. Golay FEC coding examples. 75FIGURE A-10. Word bit coding and interleaving. 77FIGURE A-11. Bit and word decoding. 78FIGURE A-12. ALE basic word structure. 80FIGURE A-13. Basic 38 subset (unshaded areas). 85FIGURE A-14. Valid word sequences. 94FIGURE A-15. Calling cycle sequence. 96FIGURE A-16. Message sequence. 99FIGURE A-17. Conclusion (terminator) sequences. 101FIGURE A-18. Valid word sequence (calling cycle section). 104FIGURE A-19. Valid word sequence (message section). 105FIGURE A-20. Valid word sequence (conclusion section). 106FIGURE A-21. Basic frame structure examples. 107FIGURE A-22. Basic sounding structure. 111FIGURE A-23. Call rejection scanning sounding protocol. 112FIGURE A-24. Call acceptance scanning sounding protocol. 113FIGURE A-25. Scanning sounding with optional handshake protocol. 116FIGURE A-26. Local noise report (optional). 121FIGURE A-27. LQA memory example. 122FIGURE A-28. Link establishment states. 125FIGURE A-29. Individual calls. 129FIGURE A-30. Response frame. 131FIGURE A-31. Acknowledgment frame. 132FIGURE A-32. Slotted responses. 135FIGURE A-33. 2G ALE slotted responses. 136FIGURE A-34. Net call. 136FIGURE A-35. Group call. 137FIGURE A-36. Power control CMD format. 147TABLE OF CONTENTS(continued)PARAGRAPH PAGEFIGURE A-37. Frequency select CMD format. 148FIGURE A-38. Time exchange CMD word. 155FIGURE A-39. Coarse time and authentication words. 156FIGURE A-40. Mode control CMD format. 157FIGURE A-41. Modem selection CMD format. 158FIGURE A-42. Crypto selection CMD format. 160FIGURE A-43. Version CMD format. 160FIGURE A-44. Capabilities query CMD format. 162FIGURE A-45. Capabilities report CMD and DATA format. 162FIGURE A-46. Expanded 64 ASCII subset (shown unshaded). 168FIGURE A-47. DTM structure example. 172FIGURE A-48. Data test message reconstruction (overlay). 177FIGURE A-49. Data test message structure and ARQ example. 184FIGURE A-50. DBM interleaver and deinterleaver. 185FIGURE A-51. DBM example. 186FIGURE A-52. AQC-ALE data exchange word. 195FIGURE A-53. Example of unit call format. 207FIGURE A-54. Example of StarNet format. 208FIGURE A-55. Example AllCall frame format. 209FIGURE A-56. Example AnyCall frame formats. 209FIGURE A-57. Example sounding frame format. 210FIGURE A-58. Example inlink transaction TRW sequences. 211FIGURE A-59. Generalized AQC-ALE control message format. 214FIGURE A-60. AQC-ALE dictionary lookup message. 215FIGURE A-61. Channel definition and meet-me function. 217FIGURE A-62. AQC-ALE slot assignment. 218FIGURE A-63. List content of database. 218FIGURE A-64. Set database activation time. 219FIGURE A-65. Define database content. 219

ANNEX A. DEFINITIONS OF TIMING SYMBOLS 221ANNEX B. TIMING 224ANNEX C. SUMMARY OF ALE SIGNAL PARAMETERS 234

A.1 GENERAL.

A.1.1 Scope.

This appendix provides details of the prescribed waveform, signal structures, protocols, and performance requirements for the second generation (2G) automatic link establishment (ALE) system.

A.1.2 Applicability.

This appendix is a mandatory part of MIL-STD-188-141 whenever ALE is a requirement to be implemented into the high frequency (HF) radio system. The functional capability described herein includes automatic signaling, selective calling, automatic answering, and radio frequency (rf) scanning with link quality analysis (LQA). The capability for manual operation of the radio in order to conduct communications with existing, older generation, non-automated manual radios, shall not be impaired by implementation of these automated features.

A.2 APPLICABLE DOCUMENTS.

A.2.1 General.

The documents listed in this section are specified in A.3, A.4, and A.5 of this standard. This section does not include documents cited in other sections of this standard or recommended for additional information or as examples. While every effort has been made to ensure the completeness of this list, document users are cautioned that they must meet all specified requirements documents cited in A.3, A.4, and A.5 of this standard, whether or not they are listed.

A.2.2 Government documents.

A.2.2.1 Specifications, standards, and handbooks.

The following specifications, standards, and handbooks form a part of this document to the extent specified herein. Unless otherwise specified, the issues of these documents are those listed in the issue of the Department of Defense Index of Specifications and Standards (DODISS) and supplement thereto, cited in the solicitation.

STANDARDS
	FEDERAL
		Federal Information Processing Standards
			FIPS PUB 1-1	Publication Code: for Information Interchange

		FEDERAL STANDARDS
		FED-STD-1003	Telecommunications: Synchronous Bit Orientation Data Link Control Procedures (Advanced Data Communications Control Procedures)
		FED-STD-1037	Telecommunications: Glossary of Telecommunications Terms
	DEPARTMENT OF DEFENSE
		MIL-STD-188-110		Interoperability and Performance Standards for HF Data Modems

(Copies of Federal Information Processing Standards (FIPS) are available at Standardization Document Order Desk, 700 Robbins Avenue, Building #4, Section D, Philadelphia, PA 19111-5094. Non-Department of Defense (DoD) users must request copies of FIPS from the National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161-2171.)

A.2.3 Non-Government publications.

The following documents form a part of this appendix to the extent specified:

INTERNATIONAL STANDARDIZATION DOCUMENTS
	North Atlantic Treaty Organization (NATO) Standardization Agreements (STANAGs)
		STANAG 4285	Characteristics of 1200/2400/3600 bps Single Tone Modems for HF Radio Links
		STANAG 4529	Characteristics of Single Tone Modulators/Demodulators for Maritime HF Radio Links with 1240 Hz Bandwidth
	International Telecommunications Union (ITU), Radio Regulations
		ITU-R F.520-2	Recommendation for Fixed Service, use of High Frequency Ionospheric Channel Simulators

(Application for copies should be addressed to the General Secretariat, International Organization for Standardization (ISO) 1, Rue de Varembe, CH-1211 Geneva 20, Switzerland.)

Other Publications
	NMSU-EE-CD-001		Wireless Network Waveform Samples

(Application for copies should be addressed to New Mexico State University, Klipsch School of Electrical and Computer Engineering, University Park, NM 88003, Attn: Dr. E. E. Johnson.)

(Non-Government standards and other publications are normally available from the organizations that prepare or distribute the documents. These documents also may be available in or through libraries or other informational services.)

A.3 DEFINITIONS.

A.3.1 Terms.

Definitions of terms used in this document shall be as specified in the current edition of
FED-STD-1037 except where inconsistent with the use in this standard. In addition, the following definitions are applicable for the purpose of this standard.

• Available State. An ALE controller is in the available state when it does not currently have a link with any other station, and is not in the process of establishing a link. An ALE controller that is programmed for multichannel scanning operation will be scanning when it is in the available state. Single-channel controllers will remain tuned to the assigned channel regardless of their state.
• Exclusive OR.. Used as a check, the condition that exits when each resulting bit is a "1" if the two input bits do not match, or the resulting bit is a "0" when the two input bits match.
• Linking State. An ALE controller enters the linking state from the available state when it sends or receives an ALE call frame. Scanning controllers stop scanning when they enter the linking state. An ALE controller returns to the available state if the linking attempt does not complete successfully. Upon successful completion of a three-way handshake, controllers in the linking state enter the linked state.
• Linked State. An ALE controller is considered to be in the linked state if it has successfully completed link establishment with one or more stations, and at least one link to which it is party has not been terminated. While in the linked state, a wait-for-activity timer will be running (if not disabled by the operator). Controllers programmed to scan will not be scanning while in the linked state. After link establishment, communication among linked stations normally is carried by additional three-way handshakes, but controllers remain in the linked state during these handshakes.

A.3.2 Abbreviations and acronyms.

The abbreviations and acronyms used in this document are defined below. Those listed in the current edition of FED-STD-1037 have been included for the convenience of the reader.

	2G ALE	second generation automatic link establishment
	3G ALE	second generation automatic link establishment

ACK

acknowledge character

AQC

Alternative Quick Call

AQC-ALE

Alternative Quick Call Automatic Link Establishment

AGC

automatic gain control

ALE

automatic link establishment

AMD

automatic message display

ARQ

automatic repeat request

ASCII

American Standard Code for Information Interchange

AWGN

Additive white gaussian noise

b/s

bits per second

BCD

binary coded decimal

BER

bit error ratio

chps

channels per second

CCIR

International Radio Consultative Committee

CMD

ALE preamble word COMMAND

CRC

cyclic redundancy check

dB

Decibel

dBw

dB referred to 1 W (watt)

DBM

data block message

DC

data code

DCE

data circuit-terminating equipment

DO

design objective

DTE

data terminal equipment

DTM

data text message

e.g.

for example

FCS

frame check sequence

FEC

forward error correction

FSK

frequency shift keying

Hz

hertz

HF

high frequency

HFNC

high frequency node controller

ID

identification

IFF

if and only if

ISDN

Integrated Services Digital Network

kHz

Kilohertz

LP

linking protection

LQA

link quality analysis

LSB

(1) lower sideband (2) least significant bit

ms

millisecond

NATO

North Atlantic Treaty Organization

ITU

International Telecommunications Union

ISO

Organization of Standardization

FIPS

Federal Information Processing Standards

DODISS

Department of Defense Index of Specifications and Standards

DoD

Department of Defense

AQC

Alternative Quick Call

MF

medium frequency

MHz

megahertz

MP

multipath

MSB

most significant bit

NAK

negative-acknowledge character

NT

Not Tested

PL

probability of linking

PPM

parts per million

rf

radio frequency

REP

ALE preamble word REPEAT

RX

receive

s

second

SCTY

Security

	SINAD	signal-plus-noise-plus-distortion to noise-plus-distortion ratio
	SN	Slot Number

SNR

signal to noise ratio

SPS

symbols per second

SSB

single-sideband [transmission]

TDMA

time-division multiple access

TIS

ALE preamble word THIS IS

TOD

time of day

TWAS

ALE preamble word THIS WAS

TX

transmit

UI

unique index

UUF

user unique function

UUT

units under test

	WRTT	wait for response and tune timeout
	WS	AQC-ALE Word Sync word

A.3.3 Definitions of timing symbols.

The abbreviations and acronyms used for timing symbols are contained in annex A to this appendix.

A.4 GENERAL REQUIREMENTS.

A.4.1 ALE introduction.

The techniques specified in this appendix employ a robust modem and forward error correction coding and constitutes a digital ALE data link. The exchange of such ALE words according to the specified protocols supports channel evaluation, selective calling, and passing data messages and constitute an ALE data link layer. (The ALE modem, radio, coupler, antenna, and so on constitute the corresponding physical layer.)

The ALE data link layer contains three sublayers, as shown in figure A-1: a lower sublayer concerned with error correction and detection (forward error correction [FEC] sublayer), an upper sublayer containing the ALE protocol (ALE sublayer), and a linking protection (LP) sublayer between. Within the FEC sublayer are redundancy and majority voting, interleaving, and Golay coding applied to the 24-bit ALE words which constitute the (FEC sublayer) service-data-unit, in terms of the Seven Layer Reference Model. The ALE sublayer specifies protocols for link establishment, data communication, and rudimentary LQA based on the capability of exchanging ALE words. The shaded area of figure A-1 indicates the contents of this appendix.

The following paragraphs specify the general requirements for ALE operation.

A.4.1.1 ALE addresses.

Stations designed to this appendix shall employ the addressing structure specified in A.5.2.4 to identify individual stations and collections of stations (nets and groups).

A.4.1.2 Scanning.

The radio system shall be capable of repeatedly scanning selected channels stored in memory (in the radio or controller) under either manual control or under the direction of any associated automated controller. The radio shall stop scanning and wait on the most recent channel upon the occurrence of any of the following selectable events:

• Automatic controller decision to stop scan (the normal mode of operation)
• Manual input of stop scan
• Activation of external stop-scan line (if provided)

The scanned channels should be selectable by groups (often called "scan lists") and also individually within the groups, to enable flexibility in channel and network scan management.

A.4.1.3 Calling.

Upon request by the operator or an external automated controller, the radio system shall execute the appropriate calling protocol specified in A.5.5.

A.4.1.4 Channel evaluation.

The radio system shall be capable of automatically transmitting ALE sounding transmissions in accordance with A.5.3, and shall automatically measure the signal quality of ALE receptions in accordance with A.5.4.1.

A.4.1.5 Channel quality display.

If an operator display is provided, the display shall have a uniform scale, 0-30 with 31 being unknown all based on signal-plus-noise-plus-distortion to noise-plus-distortion (SINAD).

A.4.2 System performance requirements.

Stations designed to this appendix shall demonstrate an overall system performance equal to or exceeding the following requirements.

A.4.2.1 Scanning rate.

Stations designed to this appendix shall incorporate selectable scan rates of two and five channels per second, and may also incorporate other scan rates (design objective (DO): 10 channels per second).

A.4.2.1.1 Alternative Quick Call (AQC) (NT).

In the optional AQC-ALE protocol, the system shall be capable of variable dwell rates while scanning such that traffic can be detected in accordance with table A-II Probability of Linking.

A.4.2.1.2 Recommendation.

Radios equipped with the optional AQC-ALE shall provide scanning at scan rates of two channels per second or five channels per second for backward compatibility to non-AQC-ALE networks.

A.4.2.2 Occupancy detection - not tested (NT).

Stations designed to this appendix shall achieve at least the following probability of detecting the specified waveforms (See A.5.4.7) under the indicated conditions, with false alarm rates of no more than 1 percent. The channel simulator shall provide additive white gaussian noise (AWGN) without fading or multipath (MP). See table A-I.

A.4.2.3 Linking probability.

Linking attempts made with a test setup configured as shown in figure A-3, using the specified ALE signal created in accordance with this appendix, shall produce a probability of linking as shown in table A-II.

The receive audio input to the ALE controller shall be used to simulate the three channel conditions. The modified International Radio Consultative Committee (CCIR) good channel shall be characterized as having 0.52 millisecond (ms) (modified from 0.50ms) MP delay and a fading (two sigma) bandwidth of 0.1 hertz (Hz). The modified CCIR poor channel, normally characterized as consisting of a circuit having 2.0 ms MP delay with a fading (two sigma) bandwidth of 1.0 Hz, shall be modified to have 2.2 ms MP delay and a fading (two sigma) bandwidth of 1.0 Hz. Doppler shifts of +60 Hz shall produce no more than a 1.0 decibel (dB) performance degradation from the requirements of table A-II for the modified CCIR good and poor channels.

NOTE: This modification is necessary due to the fact that the constant 2-ms MP delay (an unrealistic fixed condition) of the CCIR poor channel results in a constant nulling of certain tones of the ALE tone library. Other tone libraries would also have some particular MP value, which would result in continuous tone cancellation during simulator testing.

Each of the signal-to-noise (SNR) ratio values shall be measured in a nominal 3-kiloHertz (kHz) bandwidth. Performance tests of this capability shall be conducted in accordance with ITU-R F.520-2 Use of High Frequency Ionospheric Channel Simulators employing the C.C. Watterson Model. This test shall use the individual scanning calling protocol described in A.5.5.3. The time for performance of each link attempt shall be measured from the initiation of the calling transmission until the successful establishment of the link. Performance testing shall include the following additional criteria:

a. The protocol used shall be the individual scanning calling protocol with only TO and TIS preambles.

b. Addresses used shall be alphanumeric, one word (three characters) in length from the 38-character basic American Standard Code for Information Interchange (ASCII) subset.

c. Units under test (UUTs) shall be scanning 10 channels at two channels per second, and repeated at five channels per seconds.

d. Call initiation shall be performed with the UUT transmitter stopped and tuned to the calling frequency.

e. Maximum time from call initiation (measured from the start of UUT rf transmission -- not from activation of the ALE protocol) to link establishment shall not exceed 14.000 seconds, plus simulator delay time. The call shall not exceed 23 redundent words, the response three redundent words and the acknowledgment three redundent words. (See A.5.2.2.4 and Annex A).

NOTE: Performance at the higher scan rates shall also meet the foregoing requirements and shall meet or exceed the probability of linking as shown in table A-II.

A.4.2.3.1 AQC-ALE linking probability.

When the optional AQC-ALE protocol is implemented, the probability of linking shall conform to table A-II with the following additional criteria:

a. The protocol used shall be quick AQC individual calling protocol with no message passing.

b. Addresses shall be one to six characters in the 38-character basic ASCII subset.

c. Units being called shall be scanning 10 channels.

d. Call initiation shall be performed with the UUT transmitter stopped and tuned to the calling frequency.

e. The initial call probe shall not exceed 10 T_rw, the call response shall not exceed 4T_rw, and the acknowledgment shall not exceed 2 T_rw.

A.4.2.3.2 AQC-ALE linking performance.

AQC-ALE linking performance shall not be degraded in LP level 1 or 2. Scan rates of two or five channels per second may degrade performance because insufficient redundent words are emitted during the call probe.

A.4.3 Required data structures.

A.4.3.1 Channel memory.

The equipment shall be capable of storing, retrieving, and employing at least 100 different sets of information concerning channel data to include receive and transmit frequencies with associated mode information. See table A-III. The channel data storage shall be nonvolatile.

The mode information normally includes:

• transmit power level
• traffic or channel use (voice, data, etc.)
• sounding data
• modulation type (associated with frequency)
• transmit/receive modes
• filter width (DO)
• automatic gain control (AGC) setting (DO)
• input/output antenna port selection (DO)
• input/output information port selection (DO)
• noise blanker setting (DO)
• security (DO)
• sounding self address(es) SA....n(DO)

Any channel (a) shall be capable of being recalled manually or under the direction of any associated automated controller, and (b) shall be capable of having its information altered after recall without affecting the original stored information settings.

A.4.3.2 Self address memory.

The radio shall be capable of storing, retrieving, and employing at least 20 different sets of information concerning self addressing. The self-address information storage shall be nonvolatile. These sets of information include self (its own personal) address(es), valid channels which are associated for use, and net addressing. Net addressing information shall include (for each "net member" self address, as necessary) the net address and the associated slot wait time (in multiples of T_w). See table A-IV and A.5.5.4.1. The slot wait time values are T_swt(slot number (SN)) from the formula, T_swt (SN) = T_sw x SN . Stations called by their net call address shall respond with their associated self (net member) address with the specified delay (T_swt(SN)). For example, the call is "GUY," thus the response is "BEN." Stations called individually by one of their self addresses (even if a net member address) shall respond immediately and with that address, as specified in the individual scanning calling protocol. Stations called by one of their self addresses (even if a net member address) within a group call shall respond in the derived slot, and with that address, as specified in the star group scanning protocol. If a station is called by one of its net addresses and has no associated net member address, it shall pause and listen but shall not respond (unless subsequently called separately with an available self or net member address), but shall enter the linked state.

A.4.3.3 Other station table.

The radio shall be capable of storing, retrieving, and employing at least 100 different sets of information concerning the addresses of other stations and nets, channel quality data to those stations and nets (measurements or predictions), and equipment settings specific to links with each station or net.

DO: any excess capacity which is not programmed with preplanned other station information should be automatically filled with any addresses heard on any of the scanned or monitored channels. When the excess capacity is filled, it should be kept current by replacing the oldest heard addresses with the latest ones heard. This information should be used for call initiation to stations (if needed), and for activity evaluation.

A.4.3.3.1 Other station address storage.

Individual station addresses shall be stored in distinct table entries, and shall be associated with a specific wait for reply time (T_wr) if not the default value. Net information shall include own net and net member associations, relative slot sequences, and own net wait for reply times (T_wrn) for use when calling. See figure A-4. The storage for addresses and settings shall be nonvolatile.

A.4.3.3.2 Link quality memory.

The equipment shall be capable of storing, retrieving, and employing at least 4000 (DO: 10,000) sets of connectivity and LQA information associated with the channels and the other addresses in an LQA memory. The connectivity and LQA information storage shall be retained in memory for not less than one hour during power down or loss of primary power. The information in each address/channel "cell" shall include as a minimum, bilateral SINAD values of (a) the signals received at the station, and (b) the station's signals received at, and reported by, the other station. It shall also include either an indicator of the age of the information (for discounting old data), or an algorithm for automatically reducing the weight of data with time, to compensate for changing propagation conditions. (DO: the cells of the LQA memory should also include bilateral bit-error ratio (BER) and bilateral MP information derived by suitably equipped units.) The information within the LQA memory shall be used to select channels and manage networks as stated in this document. See figure A-4.

A.4.3.3.3 Other station settings storage.

DO: Equipment settings for use in linking with specific stations or nets should be stored in nonvolatile memory. Such settings may include antenna selection and azimuth, channels authorized for that station or net, power limits for the relevant net, and so on.

A.4.3.4 Operating parameters.

The following ALE operating parameters shall be programmable by the operator or an external automated controller. Complete definitions of the parameters are provided in Appendix H.

ScanRate	RequestLQA	OtherAddr	LqaStatus
MaxScanChan	AutoPowerAdj	OtherAddrStatus	LqaAge
MaxTuneTime	SelfAddrTable	OtherAddrNetMembers	LqaMultipath
TurnAroundTime	SelfAddrEntry	OtherAddrValidChannels	LqaSINAD
ActivityTimeout	SelfAddr	OtherAddrAnt	LqaBER
ListenTime	SelfAddrStatus	OtherAddrAntAzimuth	ScanSet
AcceptAnyCall	NetAddr	OtherAddrPower	ConnectionTable
AcceptAllcall	SlotWaitTime	LqaMatrix	ConnectionEntry
AcceptAMD	SelfAddrValidChannels	LqaEntry	ConnectedAddr
AcceptDTM	OtherAddrTable	LqaAddr	ConnectionStatus
AcceptDBM	OtherAddrEntry	LqaChannel

A.4.3.5 Message memory.

Storage for preprogrammed, operator entered, and incoming messages shall be provided in the equipment. This storage shall be retained in memory for not less than one hour during power down or loss of primary power. Storage for at least 12 messages (DO: 100 messages), and a total capacity of at least 1000 characters (DO: 10,000 characters) shall be provided.

A.4.4 ALE operational rules.

The ALE system shall incorporate the basic operational rules listed in table A-V. Some of these rules may not be applicable in certain applications. For example, "always listening" is not possible while transmitting with a transceiver or when using a common antenna with a separate transmitter and receiver.

A.4.5 Alternate Quick Call ALE (AQC-ALE) (NT).

A.4.5.1 Introduction.

This feature may be implemented in addition to the basic ALE functionality described in this appendix. The AQC-ALE provides a link establishment technique that requires significantly less time to link than the baseline ALE system. This is accomplished by some additional technology and trading-off some of the lesser used functions of the baseline system, for a faster linking process. The AQC-ALE shall always be listening for the baseline ALE call and shall automatically respond and operate in that mode when called.

A.4.5.2 General signaling strategies.

The AQC-ALE format employs the following characteristics:

a. Packs three address characters (21 bits) into a 16-bit value

b. Addresses are reduced from a maximum of 15 characters to 6 characters

c. Six (6) address characters are sent in every transaction

d. Replaces two seldom used preambles as follows:

• FROM preamble becomes PART2 indicating the 2nd address word
• THRU preamble becomes INLINK indicating a linked transaction

e. Isolates station addresses from message portion of the signaling structure:

• TO, TIS, TWAS, INLINK, PART2 preambles used for addressing
• CMD, DATA, and REP are used for messaging

f. Easy separation of second generation basic ALE and AQC-ALE protocols:

• Fixes 1 bit of any address word
• Prevents legitimate addresses in AQC-ALE from being legitimate addresses in second generation basic ALE.

g. Provides at least eight information bits per transmission

A.4.5.3 Features supported by AQC-ALE.

The following basic ALE features are fully implemented using the AQC-ALE protocol.

NOTE: A station operating in AQC-ALE can respond to any call type, but a station equipped with only second generation basic ALE will not respond to AQC-ALE protocol forms.

a. Linking protection levels 0, 1, 2, 3

b. Unit calls

c. Star Net calls

d. Allcalls

e. AnyCalls

f. LQA Exchange as part of the call handshake

g. Supports Orderwire and Relay features while in a link:

• automatic message display (AMD), data text message (DTM) or DBM
• User Unique Functions (UUF) when in a link
• Call Relay features
• Time of day and Network Management

h. Sound:

• Sounds are shortened to include scan time + 50percent
• Sounds may include a PSK signal to enhance LQA data

A.4.5.4 Features not provided by AQC-ALE.

a. Group call. As an alternative, a controller can use the calling protocol to add on additional members. Behavior of the system is more akin to setting up a call and then conferencing in a third party.

b. AMD, DTM, DBM are not provided during link set up. Primary focus of AQC-ALE is to establish a link between two or more stations as rapidly as possible. Once linked, information can be exchanged in the most efficient manner as is common between stations.

c. Early identification of transmitter's address during orderwire traffic or additional addressing identification for relay addresses. The need for this is eliminated because the call setup is significantly reduced. Orderwire messages are not allowed during the call setup.

A.5. DETAILED REQUIREMENTS.

A.5.1 ALE modem waveform.

A.5.1.1 Introduction.

The ALE waveform is designed to pass through the audio passband of standard SSB radio equipment. This waveform shall provide for a robust, low-speed, digital modem capability used for multiple purposes to include selective calling and data transmission. This section defines the waveform including the tones, their meanings, the timing and rates, and their accuracy.

A.5.1.2 Tones.

The waveform shall be an 8-ary frequency shift-keying (FSK) modulation with eight orthogonal tones, one tone (or symbol) at a time. Each tone shall represent three bits of data as follows (least significant bit (LSB) to the right):

• 750 Hz 000
• 1000 Hz 001
• 1250 Hz 011
• 1500 Hz 010
• 1750 Hz 110
• 2000 Hz 111
• 2250 Hz 101
• 2500 Hz 100

The transmitted bits shall be encoded and interleaved data bits constituting a word, as described in paragraphs A.5.2.2 and A.5.2.3. The transitions between tones shall be phase continuous and shall be at waveform maxima or minima (slope zero).

A.5.1.3 Timing.

The tones shall be transmitted at a rate of 125 tones (symbols) per second, with a resultant period of 8 ms per tone. Figure A-5 shows the frequency and time relationships. The transmitted bit rate shall be 375 bits per second (b/s). The transitions between adjacent redundant (tripled) transmitted words shall coincide with the transitions between tones, resulting in an integral
49 symbols (or tones) per redundant (tripled) word. The resultant single word period (T_w) shall be 130.66... ms (or 16.33... symbols), and the triple word (basic redundant format) period (3 T_w) shall be 392 ms.

A.5.1.4 Accuracy.

At baseband audio, the generated tones shall be within +1.0 Hz. At rf, all transmitted tones shall be within the range of 2.0 dB in amplitude. Transmitted symbol timing, and therefore, the bit and word rates shall be within ten parts per million.

A.5.2 Signal structure.

A.5.2.1 Introduction.

This section provides definition of the ALE signal structure. Included are: forward error correction, word structure, addressing, frame structure, and synchronization. Also described in this section are: addressing, signal quality analysis, and the functions of the standard word preambles associated with the signal structure.

A.5.2.2 FEC.

A.5.2.2.1 General.

The effective performance of stations, while communicating over adverse rf channels, relies on the combined use of forward error correction, interleaving, and redundancy. These functions shall be performed within the transmit encoder and receive decoder.

A.5.2.2.2 Golay coding.

The generator matrix G, derived from g(x), shall contain an identity matrix I₁₂ and a parity matrix P as shown in figure A-6. The corresponding parity check matrix H shall contain a transposed matrix p^T and an identity matrix I₁₂ as shown in figure A-7.

A.5.2.2.2.1 Encoding.

A.5.2.2.2.2 Decoding.

where y = x + e is a received vector which is the modulo-2 sum of a code word x and an error vector e, s is a vector of "n - k" bits called the syndrome. See figure A-9. See figure A-7 for the value of H. Each correctable/detectable error vector e results in a unique vector s. Because of this, s is computed according to the equation above and is used to index a look-up of the corresponding e, which is then added modulo-2 to y to give the original code word x. Flags are set according to the number of errors being corrected. The uses of the flags are described in A.5.2.6. If s is not equal to 0 and e contains more ones than the number of errors being corrected by decoding mode, a detected error is indicated and the appropriate flag is set.

		I12					P
	100	000	000	000	:	101	011	100	011
	010	000	000	000	:	111	110	010	010
	001	000	000	000	:	110	100	101	011
	000	100	000	000	:	110	001	110	110
	000	010	000	000	:	110	011	011	001
	000	001	000	000	:	011	001	101	101
G=	000	000	100	000	:	001	100	110	111
	000	000	010	000	:	101	101	111	000
	000	000	001	000	:	010	110	111	100
	000	000	000	100	:	001	011	011	110
	000	000	000	010	:	101	110	001	101
	000	000	000	001	:	010	111	000	111

A.5.2.2.3 Interleaving and deinterleaving.

A.5.2.2.4 Redundant words.

Each of the transmitted 49-bit (or 16-1/3 symbol) (T_w) words shall be sent redundantly (times 3) to reduce the effects of fading, interference, and noise. An individual (or net) routing word (TO...), used for calling a scanning (multichannel) station (or net), shall be sent redundantly as long as required in the scan call (T_sc) to ensure receipt, as described in A.5.5.2. However, when the call is a non-net call to multiple scanning stations (a group call, using THRU and REPEAT (REP) alternately), the first individual routing word (THRU) and all the subsequent individual routing words (REP, THRU, REP,...) shall be sent three adjacent times (T_rw). These triple words for the individual stations shall be rotated in group sequence as described in A.5.5.3. See figure A-11. At bit time intervals (approximately T_w/49), the receiver shall examine the present bit and past bit stream and perform a 2/3 majority vote, on a bit-by-bit basis, over a span of three words. See tables A-VI and A-VII. The resultant 48 (ignoring the 49th bit) most recent majority bits constitute the latest majority word and shall be delivered to the deinterleaver and FEC decoder. In addition, the number of unanimous votes of the 48 possible votes associated with this majority word are temporarily retained for use as described in A.5.2.6.

A.5.2.3 Word structures.

A.5.2.3.1 ALE word format.

The basic ALE word shall consist of 24 bits of information, designated W1 (MSB) through W24 (LSB). The bits shall be designated as shown in figure A-12.

A.5.2.3.1.1 Structure.

A.5.2.3.1.2 Word types.

The leading three bits, W1 through W3, are designated preamble bits P3 through P1, respectively. These preamble bits shall be used to identify one of eight possible word types.

A.5.2.3.1.3 Preambles.

Optional AQC-ALE preambles are defined in A.5.8.1.2.

A.5.2.3.2 Address words.

A.5.2.3.2.1 TO.

A.5.2.3.2.2 THIS IS (TIS).

The TIS word (101) shall be used as a routing designator which shall indicate the address of the present calling (or sounding) station which is directly transmitting the call (or sound). Except for the use of TWAS, TIS shall be used in all ALE protocols to terminate the ALE frame and transmission. It shall indicate the continuation of the protocol or handshake, and shall direct, request, or invite (depending on the specific protocol) responses or acknowledgments from other called or receiving stations. The TIS shall be used to designate the call acceptance sound. The TIS word itself shall contain the first three characters of the calling stations address. For extended addresses, the additional address words (and characters) shall be contained in alternating DATA and REP words which shall immediately follow, exactly as described for whole addresses using the TO word and sequence. The entire address (and the required portion of the TIS, DATA, REP, DATA, REP sequence, as necessary) shall be used only in the conclusion section of the ALE frame (or shall constitute an entire sound). TWAS shall not be used in the same frame as TIS, as they are mutually exclusive.

A.5.2.3.2.3 THIS WAS (TWAS).

A.5.2.3.2.4 THRU.

The THRU word (001) shall be used in the scanning call section of the calling cycle only with group call protocols. The THRU word shall be used alternately with REP, as routing designators, to indicate the address first word of stations that are to be directly called. Each address first word shall be limited to one basic address word (three characters) in length. A maximum of five different address first words shall be permitted in a group call. The sequence shall only be alternations of THRU, REP. The THRU shall not be used for extended addresses, as it will not be used within the leading call section of the calling cycle. When the leading call starts in the group call, the entire group of called stations shall be called with their whole addresses, which shall be sent using the TO preambles and structures, as described in A.5.2.3.2.1.

NOTE: 1. The THRU word is also reserved for future implementation of indirect and relay protocols, in which cases it may be used elsewhere in the ALE frame and with whole addresses and other information. Stations designed in compliance with this nonrelay standard should ignore calls to them which employ their address in a THRU word in other than the scanning call.

NOTE: 2. The THRU preamble value is also reserved for the AQC-ALE protocol.

A.5.2.3.2.5 FROM.

NOTE: 1. The FROM word is also reserved for future implementation of indirect and relay protocols, in which cases it may be used elsewhere in the ALE frame and with multiple addresses and other information. Stations designed in compliance with this nonrelay standard should ignore sections of calls to them that employ FROM words in any other sequence than immediately before the CMD word.

NOTE: 2. The FROM preamble value is also reserved for the AQC-ALE protocol.

A.5.2.3.3 Message words.

All message words (orderwire messages) begin with a word with the CMD preamble.

A.5.2.3.3.1 CMD.

The CMD word (110) is a special orderwire designator which shall be used for system-wide coordination, command, control, status, information, interoperation, and other special purposes. CMD shall be used in any combination between ALE stations and operators. CMD is an optional designator which is used only within the message section of the ALE frame, and it shall have (at some time in the frame) a preceding call and a following conclusion, to ensure designation of the intended receivers and identification of the sender. The first CMD terminates the calling cycle and indicates the start of the message section of the ALE frame. The orderwire functions are directed with the CMD itself, or when combined with the REP and DATA words. See A.5.6 for message words (orderwire messages) and functions.

A.5.2.3.4 Extension words.

A.5.2.3.4.1 DATA.

The DATA word (000) is a special designator which shall be used to extend the data field of any previous word type (except DATA itself) or to convey information in a message. When used with the routing designators TO, FROM, TIS, or TWAS, DATA shall perform address extension from the basic three characters to six, nine, or more (in multiples of three) when alternated with REP words. The selected limit for address extension is a total of 15 characters. When used with CMD, its function is predefined as specified in A.5.6 for message words (orderwire messages) and functions.

A.5.2.3.4.2 REP.

The REP word (111) is a special designator which shall be used to duplicate any previous preamble function or word meaning while changing the data field contents (bits W4 through W24). See table A-VIII. Any change of words or data field bits requires a change of preamble bits (P₃ through P₁) to preclude uncertainty and errors. If a word is to change, even if the data field is identical to that in the previous word, the preamble shall be changed, thereby clearly designating a word change. When used with the routing designator TO, REP performs address expansion, which enables more than one address to be specified. See A.5.2.3.2.4 for use with THRU. With DATA, REP may be used to extend and expand address, message, command, and status fields. REP shall be used to perform these functions, and it may directly follow any other word type except for itself, and except for TIS or TWAS, as there cannot be more than one transmitter for a specific call at a given time.

NOTE 1. REP is used in T_sc of group calls directed to units with different first word addresses.

NOTE 2. REP is not used in T_sc of calls directed to groups with same first word addresses. Also REP is not used in T_sc of calls directed to individuals and nets.

A.5.2.4 Addressing.

A.5.2.4.1 Introduction.

The ALE system deploys a digital addressing structure based upon the standard 24-bit (three character) word and the Basic 38 character subset. As described below, ALE stations have the capability and flexibility to link or network with one or many prearranged or as-needed single or multiple stations. All ALE stations shall have the capacity to store and use at least 20 self addresses of up to 15 characters each in any combination of individual and net calls. There are three basic addressing methods which will be presented:

• Individual station
• Multiple station
• Special modes

NOTE: Certain alphanumeric address combinations may be interpreted to have special meanings for emergency or specific functions, such as "SOS," "MAYDAY," "PANPAN," "SECURITY," "ALL," "ANY," and "NULL." These should be carefully controlled or restricted.

A.5.2.4.2 Basic 38 subset.

A.5.2.4.3 Stuffing.

The ALE basic address structure is based on single words which, in themselves, provide multiples of three characters. The quantity of available addresses within the system, and the flexibility of assigning addresses, are significantly increased by the use of address character stuffing. This technique allows address lengths that are not multiples of three to be compatibly contained in the standard (multiple of three characters) address fields by "stuffing" the empty trailing positions with the utility symbol "@." See table A-IX. "Stuff-1" and "Stuff-2" words shall only be used in the last word of an address, and therefore should appear only in the leading call (T_lc) of the calling cycle (T_cc).

NOTE: As an example of proper usage, a call to the address "MIAMI" would be structured "TO MIA," "DATA MI@."

A.5.2.4.4 Individual addresses.

The fundamental address element in the ALE system is the single routing word, containing three characters, which forms the basic individual station address. This basic address word, used primarily for intranet and slotted operations, may be extended to multiple words and modified to provide increased address capacity and flexibility for internet and general use. An address which is assigned to a single station (within the known or used network) shall be termed an "individual" address. If it consists of one word (that is, no longer than three characters) it shall be termed a "basic" size, and if it exceeds one word, it shall be termed an "extended" size.

A.5.2.4.4.1 Basic size.

The basic address word shall be composed of a routing preamble (TO, or possibly a REP which follows a TO, in T_lc of group call, or a TIS or TWAS) plus three address characters, all of which shall be alphanumeric numbers of the Basic 38 subset. The three characters in the basic individual address provide a Basic 38-address capacity of 46,656, using only the 36 alphanumerics. This three-character single word is the minimum structure. In addition, all ALE stations shall associate specific timing and control information with all own addresses, such as prearranged delays for slotted net responses. As described in A.5.5, the basic individual addresses of various station(s) may be combined to implement flexible linking and networking.

NOTE: All ALE stations shall be assigned at least one (DO: several) single-word address for automatic use in one-word address protocols, such as slotted (multistation type) responses. This is a mandatory user requirement, not a design requirement. However, nothing in the design shall preclude using longer addresses.

A.5.2.4.4.2 Extended size.

Extended addresses provide address fields which are longer than one word (three characters), up to a maximum system limit of five words (15 characters). See table A-X. This 15-character capacity enables Integrated Services Digital Network (ISDN) address capability. Specifically, the ALE extended address word structure shall be composed of an initial basic address word, such as TO or TIS, as described above, plus additional words as necessary to contain the additional characters in the sequence DATA, REP, DATA, REP, for a maximum total of five words. All address characters shall be the alphanumeric members of the Basic 38 subset.

NOTE 1: All ALE stations shall be assigned at least one (DO: several) two-word address(es) for general use, plus an additional address(es) containing the station's assigned call sign(s). This is a mandatory user requirement, not a design requirement. However, nothing in the design shall preclude using longer addresses.

NOTE 2: The recommended standard address size for intranet, internet, and general non-ISDN use is two words. Any requirement to operate with address sizes larger than six characters must be a network management decision. As examples of proper usage, a call to "EDWARD" would be "TO EDW," "DATA ARD," and a call to "MISSISSIPPI" would be "TO MIS," "DATA SIS," "REP SIP," "DATA PI@."

A.5.2.4.5 Net addresses.

The purpose of a net call is to rapidly and efficiently establish contact with multiple prearranged (net) stations (simultaneously if possible) by the use of a single net address, which is an additional address assigned to all net members in common. When a net address type function is required, a calling ALE station shall use an address structure identical to the individual station address, basic or extended as necessary. For each net address at a net member's station, there shall be a response slot identifier, plus a slot width modifier if directed by the specific standard protocol. As described in paragraphs A.5.5.3 and A.5.5.4, additional information concerning the assigned response slots (and size) must be available, and the mixing of individual, net, and group addresses and calls is restricted

A.5.2.4.6 Group addresses.

The purpose of a group call is to establish contact with multiple nonprearranged (group) stations (simultaneously if possible) rapidly and efficiently by the use of a compact combination of their own addresses which are assigned individually. When a group address type function is required, a calling ALE station shall use a sequence of the actual individual station addresses of the called stations, in the manner directed by the specific standard protocol. A station's address shall not appear more than once in a group calling sequence, except as specifically permitted in the group calling protocols described in A.5.5.4.

NOTE: The group feature is not available in the AQC-ALE protocol.

A.5.2.4.7 Allcall addresses.

An "AllCall" is a general broadcast that does not request responses and does not designate any specific address. This mechanism is provided for emergencies ("HELP!"), broadcast data exchanges, and propagation and connectivity tracking. The global AllCall address is "@?@." The AllCall protocol is discussed in A.5.5.4.4. As a variation on the AllCall, the calling station can organize (or divide) the available but unspecified receiving stations into logical subsets, using a selective AllCall address. A selective AllCall is identical in structure, function, and protocol to the AllCall except that it specifies the last single character of the addresses of the desired subgroup of receiving stations (1/36 of all). By replacing the "?" with an alphanumeric, the selective AllCall special address pattern is "TO @A@" (or possibly "THRU @A@" and "REP @B@" if more than one subset is desired), where "A" (and "B," if applicable) in this notation represents any of the 36 alphanumerics in the Basic-38 subset. "A" and "B" may represent the same or different character from the subset, and specifically indicate which character(s) must be last in a station's address in order to stop scan and listen.

NOTE: For ACQ-ALE, the Part2 address portion shall contain the same three characters used in the TO word of the call.

A.5.2.4.8 AnyCalls.

An "AnyCall" is a general broadcast that requests responses without designating any specific addressee(s). It is required for emergencies, reconstitution of systems, and creation of new networks. An ALE station may use the AnyCall to generate responses from essentially

unspecified stations, and it thereby can identify new stations and connectivities. The global AnyCall address is "@@?." The AnyCall protocol is discussed in A.5.5.4.5. If too many responses are received to an AnyCall, or if the caller must organize the available but unspecified responders into logical subsets, a selective AnyCall protocol is used. The selective AnyCall address is identical in structure, function, and protocol to the global AnyCall, except that it specifies the last single character of the addresses of the desired subset of receiving station (1/36 of all). By replacing the "?" with an alphanumeric, the global AnyCall becomes a selective AnyCall whose special address pattern is "TO @@A." If even narrower acceptance and response criteria are required, the double selective AnyCall should be used. The double selective AnyCall is an operator selected general broadcast which is identical to the selective AnyCall described above, except that its special address (using "@AB" format) specifies the last two characters that the desired subset of receiving stations must have to initiate a response.

NOTE: For ACQ-ALE, the Part2 address portion shall contain the same three characters used in the TO word of the call.

A.5.2.4.9 Wildcards.

A "wildcard" is a special character that the caller uses to address multiple-station addresses with a single-call address. The receivers shall accept the wildcard character as a substitute for any alphanumeric in their self addresses in the same position or positions. Therefore, each wildcard character shall substitute for any of 36 characters (A to Z, 0 to 9) in the Basic 38-character subset. The total lengths of the calling (wildcard) address, and the called addresses shall be the same. The special wildcard character shall be "?" (0111111). It shall substitute for any alphanumeric in the Basic 38-character subset. It shall substitute for only a single-address character position in an address, per wildcard character. See table A-XI for examples of acceptable patterns.

A.5.2.4.10 Self addresses.

For self test, maintenance, and other purposes, stations shall be capable of using their own self addresses in calls. When a self-addressing type function is required, ALE stations shall use the following self-addressing structures and protocols. Any ALE calling structures and protocols permissible within this standard, and containing a specifically addressed calling cycle (such as "TO ABC," but not AllCall or AnyCall), shall be acceptable, except that the station may substitute (or add) any one (or several) of its own calling addresses into the calling cycle.

A.5.2.4.11 Null address.

For test, maintenance, buffer times, and other purposes, the station shall use a null address that is not directed to, accepted by, or responded to by any station. When an ALE station requires a null address type function, it shall use the following null address protocol. The null address special address pattern shall be "TO @@@," (or "REP @@@"), if directly after another TO. The null address shall only use the TO (or REP), and only in the calling cycle (T_cc). Null addresses may be mixed with other addresses (group call), in which case they shall appear only in the leading call (T_lc), and not in the scanning call (T_sc). Nulls shall never be used in conclusion (terminator) (TIS or TWAS). If a null address appears in a group call, no station is designated to respond in the associated slot; therefore, it remains empty (and may be used as a buffer for tune-ups, or overflow from the previous slot's responder, etc.).

A.5.2.4.12 In-link address.

The inlink address feature is used by a system to denote that all members in the established link are to act upon the information sent in the frame containing the inlink address. The inlink address shall be '?@?'. When a radio enters the linked condition with one or more stations, the radio shall expand the set of recognized self addresses to include the inlink address ('?@?'). When a frame is transmitted by any member of the link using the inlink address, all members are thus addressed publicly and are to use the frame information. Thus, if a linked member sent an AMD message, all members would present that message to their user. If the member sent a frame terminated with a TWAS preamble, then all members would note that the transmitting station just 'left' the link. Short messages of 'to-F?@?, to-?@?, tis-TALKINGMEMBER' would act as a keep-alive function and cause the receiving radio to extend any link termination timer.

A.5.2.5 Frame structure.

All ALE transmissions are based on the tones, timing, bit, and word structures described in paragraphs A.5.1 and A.5.2.3. All calls shall be composed of a "frame," which shall be constructed of contiguous redundant words in valid sequence(s) as described in figure A-14, as limited in table A-VII, and in formats as described in A.5.5. There are three basic frame sections: calling cycle, message, and conclusion. See A.5.2.5.5 for basic frame structure examples.

A.5.2.5.1 Calling cycle.

The initial section of all frames (except sounds) is termed a calling cycle (T_cc), and it is divided into two parts: a scanning call (T_sc) and a leading call (T_lc). The scanning call shall be composed of TO words if an individual or net call (or THRU and REP words, alternating, if a group call), which contain only the first word(s) of the called station(s) or net address. The leading call shall be composed of TO (and possibly DATA and REP) words containing the whole address(es) for the called station(s), from initiation of the leading call until the start of the message section or the conclusion (thus the end of the calling cycle). See figure A-15. The use of REP and DATA is described in A.5.2.4. The set of different address first words (T_cl) may be repeated as necessary for scanning calling (T_sc), to exceed the scan period (T_s). There is no unique "flag word" or "sync word" for frame synchronization (as discussed below). Therefore, stations may acquire and begin to read an ALE signal at any point after the start. The transmitter shall have reached at least 90 percent of the selected rf power within 2.5 ms of the first tone transmission following call initiation. The end of the calling cycle may be indicated by the start of the optional quick-ID, which occupies the first words in the message section, after the leading call and before the start of the rest of the message (or conclusion, if no message) section.

NOTE 1: The frame time may need to be delayed (equipment manufacturer dependent) to avoid loss of the leading words if the transmitter attack time is significantly long. Alternatively, the modem may transmit repeated duplicates of the scanning cycle (set of) first word(s) to be sent (not to be counted in the frame) as the transmitter rises to full power (and may even use the ALE signal momentarily instead of a tuning tone for the tuner), and then start the frame when the power is up.

NOTE 2: The 2.5-ms permissible delay of the first ALE tone, after the transmitter has reached 90 percent of selected power, is in addition to the allowable attack time delay specified in 5.3.5.1.

NOTE 3: Non-compliance with the 90 percent of power parameter will impact the probability of linking. Compliance testing for this can be construed to be met if the probability of linking criteria is met (see table A-I).

A.5.2.5.2 Message section.

The second and optional section of all frames (except sounds) is termed a "message." Except for the quick-ID, it shall be composed of CMD (and possibly REP and DATA) words from the end of the calling cycle until the start of the conclusion (thus the end of the message). The optional quick-ID shall be composed of FROM (and possibly REP and DATA) word(s), containing the transmitter's whole address. It shall only be used once at the start of the CMD message section sequences. The quick-ID enables prompt transmitter identification and should be used if the message section length is a concern. It is never used without a following (CMD...) message(s). The message section shall always start with the first CMD (or FROM with later CMD(s)) in the call. See figure A-16. The use of REP and DATA is described in A.5.7.3. The message section is not repeated within the call (although messages or information itself, within the message section, may be).

For AQC-ALE, the message section in AQC-ALE is available when in a link. The acknowledgement leg (third leg) of a call may be used as an inlink entry condition. See A.5.8.2.3.

A.5.2.5.3 Conclusion.

The third section of all frames is termed a "conclusion." It shall be composed of either TIS or TWAS (but not both) (and possibly DATA and REP) words, from the end of the message (or calling cycle sections, if no message) until the end of the call. See figure A-17. Sounds and exception shall start immediately with TIS (or TWAS) words as described in A.5.3. REP shall not immediately follow TIS or TWAS. Both conclusions and sounds contain the whole address of the transmitting station.

A.5.2.5.4 Valid sequences.

The eight ALE words types that have been described shall be used to construct frames and messages only as permitted in figures A-18, A-19, and A-20. The size and duration of ALE frames, and their parts, shall be limited as described in table A-XII.

A.5.2.5.5 Basic frame structure examples.

Contained in figure A-21 are basic examples (does not include the optional message section) of frame construction. Included are single-word and multiple-word examples of either single or multiple called station address(es) for non-scan (single-channel) and scanning (multiple-channel) use in individual, net, or group calls.

A.5.2.6 Synchronization.

The ALE system is inherently asynchronous and does not require any form of system synchronization, although it is compatible with such techniques. Within a frame, the imbedded timing and structure of the system provide the necessary "hooks" for achieving and maintaining word synchronization (word sync) during linking, orderwire, and anti-interference functions, as described herein.

A.5.2.6.1 Transmit word phase.

The ALE transmit modulator accepts digital data from the encoder and provides modulated baseband audio to the transmitter. The signal modulation is strictly timed as described in A.5.1.3 and A.5.1.4. After the start of the first transmission by a station, the ALE transmit modulator shall maintain a constant phase relationship, within the specified timing accuracy, among all transmitted triple redundant words at all times until the final frame in the transmission is terminated. Specifically,

T_{(later triple redundant word)} - T_{(early triple redundant word)} = n x T_rw

where T_{( )} is the event time of a given triple redundant word within any frame, T_rw is the period of three words (392 ms), and n is any integer.

NOTE: Word phase tracking will only be implemented within a transmission and not between transmissions.

The internal word phase reference of the transmit modulator shall be independent of the receiver (which tracks incoming signals) and shall be self timed (within its required accuracy). See A.5.1.4.

NOTE: In some applications, a single transmission may contain several frames.

A.5.2.6.2 Receiver word sync.

The receive demodulator accepts baseband audio from the receiver; acquires, tracks, and demodulates ALE signals; and provides the recovered digital data to the decoders. See figure
A-11. In data block message (DBM) mode, the receive demodulator shall also be capable of reading single data bits for deep deinterleaving and decoding.

A.5.2.6.3 Synchronization criteria.

The decoder accepts digital data from the receive demodulator and performs deinterleaving, decoding, FEC, and data checking. During initial and continuing synchronization, all of the following criteria should be used to discriminate and read every ALE word:

• Must meet or exceed a threshold of unanimous votes in the 2/3 majority voter decoder
• Successful Golay decode of "A" word bits
• Successful Golay decode of "B" word bits
• Acceptable preamble according to valid word sequences as shown in figure A-14
• Acceptable first character bits (of Basic 38 subset)
• Acceptable second character bits (of Basic 38 subset)
• Acceptable third character bits (of Basic 38 subset)
• History, status, expectations, and protocol
• Correct triple redundant word phase

The number of unanimous votes provides an easily adjustable BER signal quality discrimination, and the threshold should be chosen by the manufacturer to optimize performance. A successful Golay decode indicates that all detected bit errors were corrected within the power of the FEC code; that is, the errors were within correctable limits and therefore, the uncorrectable error flag(s) did not occur. The correction power (mode) of the Golay code should be chosen by the manufacturer to optimize performance using any of the four modes: (3/4, 2/5, 1/6, 0/7) where n/m indicates up to "n" errors detected and corrected, or up to "m" errors detected but not correctable. Acceptable preambles, as described here and defined in A.5.2.3.1.3, refer to those preambles which are within the limits of this standard. As a DO, automatic adjustment of the unanimous vote threshold and Golay mode should be provided to optimize performance under varying conditions.

NOTE: The application of each preamble is dependent on the recent signaling history of the stations heard, the active status of the machine, the handshake(s) expected, and the protocol being used, if any. For example, an uncommitted station, awaiting calls, would accept TO if individual or net call (and possibly THRU or REP if group call) as valid preambles for calls to it. It would reject CMD as being irrelevant (because it missed the preceding and required calling cycle T_cc). It might also reject TIS or TWAS (unless collecting sounding information). Acceptable characters means that each character is within the appropriate ASCII subset. Note that all criteria, together, must be satisfied to accept a word. For example, all three characters would have to be within the Basic 38 subset if a routing preamble such as a TO was decoded. Likewise, any bit combination would be conditionally acceptable if an initial REP was received, but in most cases, without the necessary knowledge of the previous word, it would be considered irrelevant and should be rejected.

A.5.3 Sounding.

A.5.3.1 Introduction.

The sounding signal is a unilateral, one-way transmission performed at periodic intervals on unoccupied channels. To implement, a timer is added to the controller to periodically initiate sounding signals (if the channel is clear). Sounding is not an interactive, bilateral technique, such as polling. However, the identification of connectivity from a station by hearing its sounding signal does indicate a high probability (but not guarantee) of bilateral connectivity and it may be done passively at the receiver. Sounding uses the standard ALE signaling, any station can receive sounding signals. As a minimum, the signal (address) information shall be displayed to the operator and, for stations equipped with connectivity and LQA memories, the information shall be stored and used later for linking. If a station has had recent transmissions on any channels that are to be sounded on, it may not be necessary to sound on those channels again until the sounding interval, as restarted from those last transmissions, has elapsed. In addition, if a net (or group) of stations is polled, their responses shall serve as sounding signals for the other net (or group) receiving stations. All stations shall be capable of performing periodic sounding on clear prearranged channels. The sounding capability may be selectively activated by, and the period between sounds shall be adjustable by the operator or controller, according to system requirements. When available, and not otherwise committed or directed by the operator or controller, all ALE stations shall automatically and temporarily display the addresses of all stations heard, with an operator selectable alert.

The structure of the sound is virtually identical to that of the basic call; however, the calling cycle is not needed and there is no message section. It is only necessary to send the conclusion (terminator) that identifies the transmitting station. See figure A-22. The type of word, either TIS or TWAS (but never both), indicates whether potential callers are encouraged or ignored, respectively. The minimum redundant sound time (T_rs) is equal to the standard one-word address leading call time (T_lc)=784 ms. Described below are both single-channel and multiple-channel protocols, plus detailed timing and control information, for designing stations.

A.5.3.2 Single channel.

The fundamental capability to automatically sound on a channel shall be in accordance with the sounding protocol as shown in figure A­22. As an option, stations may employ this protocol for single-channel sounding, connectivity tracking, and the broadcast of their availability for calls and traffic. The basic protocol consists of only one part: the sound. The sound contains its own address ("TIS A"). If "A" is encouraging calls and receives one, "A" shall follow the sound with the optional handshake protocol described in A.5.3.4. If "A" plans to ignore calls, it shall use the TWAS, which advises "B" and the others not to attempt calls, and then "A" shall immediately return to normal "available." In some systems it is necessary for a multichannel station "A" to periodically sound to a single-channel network, usually to inform them that he is active and available on that channel, although scanning. Upon receipt of "A's" sound, "B" (see figure A-23) and the other stations shall display "A's" address as a received sound and, if they have an LQA and connectivity memory, they shall store the connectivity information.

A.5.3.3 Multiple channels.

Sounding must be compatible with the scanning timing. All stations shall be capable of performing the scanning sounding protocols described herein, even if operating on a fixed frequency. See figures A-22, A-23, and A-24. These protocols establish and positively confirm unilateral connectivity between stations on any available mutually scanned channel, and they assist in establishment of links between stations waiting for contact. Stations shall employ these protocols for multichannel sounding, connectivity tracking, and the broadcast of their availability for calls and traffic.

All timing considerations and computations for individual scanning calling shall apply to scanning sounding, including sounding cycle times and (optional) handshake times.

NOTE: The scanning sound is identical to the single-channel sound except for the extension of the redundant sound time (T_rs) by adding words to the scan sounding time (T_ss) to form a scanning redundant sound time (T_srs); that is T_srs = T_ss + T_rs. The scan sounding time (T_ss) is identical in purpose to the scan calling time (T_sc) for an equivalent scanning situation, but it only uses the whole address of the transmitter.

The channel-scanning sequences and selection criteria for individual scanning calling shall also apply to scanning sounding. The channels to be sounded are termed a "sound set," and usually are identical to the "scan set" used for scanning. See figure A-23. In this illustration, station "A" is sounding and station "B" is scanning normally. If a station "A" plans to ignore calls (from "B"), which may follow "A's" sound, the following call rejection scanning sounding protocol shall be used. In a manner identical to the previously described individual scanning call, "A" lands on the first channel in the scan set (1), waits (T_wt) to see if the channel is clear (3), tunes (T_t) its coupler, comes to full power, and initiates the frame of the scanning redundant sound times (T_srs). This scanning sound is computed to exceed "B's" (and any others) scan period (T_s) by at least a redundant sound time (T_rs), which will ensure an available detection period exceeding T_drw = 784 ms. In this five-channel example, with "B" scanning at 5 chps, "A" sounds for at least 12 T_rw (4704 ms). "A" also uses "TWAS A," redundantly to indicate that calls are not invited. Upon completion of the scanning sounding frame transmission, "A" immediately leaves the channel and goes to the next channel in the sound set. This procedure repeats until all channels have been sounded, or skipped if occupied. When the calling ALE station has exhausted all the prearranged sound set channels, it shall automatically return to the normal "available" receive scan mode. As shown in figure A-23, the timing of both "A" and "B" have been prearranged to ensure that "B" has at least one opportunity, on each channel, to arrive and "capture" "A's" sound. Specifically, "B" arrives, detects sounds, waits for good words, reads at least three (redundant) "TWAS A" (in 3 to 4 T_w), stores the connectivity information (if capable), and departs immediately to resume scan.

There are several specific protocol differences when station "A" plans to welcome calls after the sound. See figure A-24. In this illustration, "A" is sounding and "B" is scanning normally. If station "A" plans to welcome calls (from "B"), which may follow his sound, the following call acceptance scanning sounding protocol shall be used. In this protocol, "A" sounds for the same time period as before. However, since "A" is receptive to calls, he shall use his normal scanning dwell time (T_d) or his preset wait before transmit time (T_wt), whichever is longer, to listen for both channel activity and calls before sounding. If the channel is clear, "A" shall initiate the scanning sound identically to before, but with "TIS A." At the end of the sounding frame, "A" shall wait for calls identically to the wait for reply and tune time (T_wrt) in the individual scanning calling protocol, in this case shown to be 6 T_w (for fast-tuning stations). During this wait, "A" shall (as always) be listening for calls that may coincidentally arrive even though unassociated with "A's" sound, plus any other sound heard, which "A" shall store as connectivity information if polling-capable. If no calls are received, "A" shall leave the channel.

A.5.3.4 Optional handshake.

In the previous descriptions, one alternative action is the implementation of an optional handshake with a station immediately after its sound. This protocol is identical in all regards to the single channel individual call protocol, except that it is manually or automatically (operator or controller) triggered by acquisition of connectivity from the station that is to be called. See figure A-25. In this illustration, "A" is scanning sounding and is receptive to calls, and "B" is receive scanning (or waiting in ambush on a channel) and requires contact with "A" if heard. "A" uses the standard call acceptance scanning sound, including the "TIS A" and the pause for calls. In this case "B" calls "A." When ALE stations are scanning sounding and receptive to calls, or required contact with such a station, the optional handshake protocol should be used. The calling station should immediately initiate the call upon the determination that the station to be called has terminated its transmission. A wait time before transmit time is not required. Therefore, if "B" hears "A's" sound and is seeking "A," "B" calls immediately using the simple single-channel call. Also, if "B's" operator or controller identifies "A's" address it can attempt the optional handshake.

A.5.4 Channel selection.

Channel selection is based on the information stored within the LQA memory (such as BER, SINAD, and MP) and this information is used to speed connectivity and to optimize the choice of quality channels. When initiating scanning (multichannel) calling attempts, the sequence of channels to be tried shall be derived from information in the LQA memory with the channel(s) with the "best score(s)" being tried first (unless otherwise directed by the operator or controller) until all the LQA scored channels are tried. However, if LQA or other such information is unavailable (or it has been exhausted and other valid channels remain available and untried) the station shall continue calling on those channels until successful or until all the remaining (untried valid) channels have been tried.

A.5.4.1 LQA.

LQA data shall be used to score the channels and to support selection of a "best" (or an acceptable) channel for calling and communication. LQA shall also be used for continual monitoring of the link(s) quality during communications that use ALE signaling. The stored values shall be available to be transmitted upon request, or as the network manager shall direct. Unless specifically and otherwise directed by the operator or controller, all ALE stations shall automatically insert the CMD LQA word (t) in the message section of their signals and handshakes when requested by the handshaking station(s), when prearranged in a network, or when specified by the protocol. See A.5.4.2. If an ALE station requires, and is capable of using LQA information (polling-capable), it may request the data from another station by setting the control bit KA1 to "1" in the CMD LQA word. If an ALE station, which is sending CMD LQA in response to a request is incapable of using such information itself (not polling-capable), it shall set the control bit KA1 to "0." It will be a network management decision to determine if the LQA is to be active or passive. For human factor considerations, LQA scores that may be presented to the operator should have higher (number) scores for better channels.

A.5.4.1.1 BER.

Analysis of the BER on rf channels, with respect to poor channels and the 8-ary modulation, plus the design and use of both redundancy and Golay FEC, shows that a coarse estimate of BER may be obtained by counting the number of non-unanimous (2/3) votes (out of 48) in the majority vote decoder. The range of this measure is 0 through 48. Correspondence to actual BER values is shown in table A-XIII.

After an ALE receiver achieves word synchronization (see A.5.2.6.2), all received words in a frame shall be measured, and a linear average BER/LQA shall be computed as follows:

• If the Golay decoder reports no uncorrectable errors in both halves of the ALE word, the number of non-unanimous votes detected in the word shall be added to the total.
• If at least one half of the ALE word contained uncorrectable errors, the number of non-unanimous votes detected shall be discarded, and 48 (the maximum value) shall be added to the total.

At the end of the transmission, the total shall be divided by the number of words received, and the total shall be stored in the Link Quality Memory as the most current BER code for the station sending the measured transmission and the channel that carried it.

A.5.4.1.2 SINAD.

The signal to noise and distortion measurement shall be a SINAD measurement ((S+N+D)/(N+D)) averaged over the duration of each received ALE signal. The SINAD values shall be measured on all ALE signals.

A.5.4.1.3 MP (optional).

Measurement of MP using received ALE signals is optional.

A.5.4.1.4 Operator display (optional).

Display of SINAD values shall be in dB.

A.5.4.2 Current channel quality report (LQA CMD).

This mandatory function is designed to support the exchange of current LQA information among ALE stations. The CMD LQA word shall be constructed as shown in table A-XIV The preamble shall be CMD (110) in bits P3 through P1 (W1 through W3). The first character shall be "a" (1100001) in bits C1-7 through C1-1 (W4 through W10), which shall identify the LQA function "analysis." It carries three types of analysis information (BER, SINAD, and MP) which are separately generated by the ALE analysis capability. Note that when the control bit KA1 (W11) is set to "1," the receiving station shall respond with an LQA report in the handshake. If KA1 is set to "0," the report is not required.

A.5.4.2.1 BER field in LQA CMD.

Measurement and reporting of BER is mandatory. The BER field in the LQA CMD shall contain five bits of information, BE5 through BE1 (W20 through W24). Refer to table A-XIII for the assigned values.

A.5.4.2.2 SINAD.

SINAD shall be reported in the CMD LQA word as follows. The SINAD is represented as five bits of information SN5 through SN1 (W15 through W19). The range is 0 to 30 dB in 1-dB steps. 00000 is 0 dB or less, and 11111 is no measurement.

A.5.4.2.3 MP.

If implemented, MP measurements shall be reported in CMD LQA words in the three bits, MP3 through MP1 (W12 through W14). The measured value in ms shall be reported rounded to the nearest integer, except that values greater than 6 ms shall be reported as 6 (110). When MP is not measured, the reported MP value shall be 7 (111).

A.5.4.3 Historical LQA report.

See MIL-STD-187-721.

A.5.4.4 Local noise report CMD (optional).

The Local Noise Report CMD provides a broadcast alternative to sounding that permits receiving stations to approximately predict the bilateral link quality for the channel carrying the report. An example application of this optional technique is networks in which most stations are silent but need to have a high probability of linking on the first attempt with a base station. A station receiving a Local Noise Report can compare the noise level at the transmitter to its own local noise level, and thereby estimate the bilateral link quality from its own LQA measurement of the received noise report transmission. The CMD reports the mean and maximum noise power measured on the channel in the past 60 minutes.

The Local Noise Report CMD shall be formatted as shown in figure A-26. Units for the Max and Mean fields are dB relative to 0.1 µV 3 KHz noise. If the local noise measurement to be reported is 0 dB or less, a 0 is sent. For measured noise ratios of 0 dB to +126 dB, the ratio in dB is rounded to an integer and sent. For noise ratios greater than +126 dB, 126 is sent. The code 127 (all 1s) is sent when no report is available for a field. By comparing the noise levels reported by a distant station on several channels, the station receiving the noise reports can select a channel for linking attempts based upon knowledge of both the propagation characteristics and the interference situation at that destination.

3	7	7	7
CMD	Noise Report (ASCII 'n')	Max	Mean
110	1101110

A.5.4.5 Single-station channel selection.

All stations shall be capable of selecting the (recent) best channel for calling or listening for a single station based on the values in the LQA memory.

A.5.4.5.1 Single-station channel selection for link establishment.

When selecting a channel for a two-way link, link quality measurements for both directions on each frequency must be considered. Figure A-27 represents a simple LQA memory example. For each address/channel cell, the measured LQA (upper section) and reported LQA values (lower section) are stored. Bilateral (handshake) scores in this example are the sum of the two LQA values.

NOTE 1: For operator viewing, LQA values for better channels should be displayed as higher numbers, and values for poorer channels should be displayed as lower numbers.

NOTE 2: In the example shown in figure A-27, if a handshake is required with station "B," channel C3 would be the best because the "round trip" (bilateral) score would be 5 (1+4), thus the lowest, channel C4 is next best with a score of 6 (3+3), the C5 with 7, C2 with 12, and C6 with 18. Linking attempts should be made in that order (C3, C4, C5, C2, and C6).

C1 is left until last because of the "x", which indicates that a recent attempted handshake on that channel failed to link. Similarly, an attempt to call "A" would yield the sequence C3(3), C5(12), C2(12), C1(24), C6(26), and C4(x). In this case, C5 was equal to C2 (both are 12), but C5 was chosen first because the paths were more balanced (LQA values were more equal).

A.5.4.5.2 Single-station channel selection for one-way broadcast.

If only a one-way transmission to a station is required instead of a handshake, the scores reported by the destination station (TO section in figure A-27) should be given greater weight than the scores measured on transmissions from that station.

NOTE: In the example, to reach "B," the sequence would be C4(3), C3(4), C5(5), C2(7), C6(12), and C1(x) as a last resort.

A.5.4.5.3 Single-station channel selection for listening.

When selecting a channel to listen for another station, the scores measured on transmissions from that station (FROM section in figure A-27) should be given greater weight than the scores reported by the destination station.

NOTE: In the example, to listen for "A," channel C4(0) would be best, and if only three channels were to be scanned, they should be C4, C3, and C2.

A.5.4.6 Multiple-station channel selection.

A station shall also be capable of selecting the (recent) best channel to call or listen for multiple stations, based on the values in the LQA memory.

NOTE: In the example shown in figure A-27, if a multiple-station handshake is required with stations "B" and "C," C5 is the best choice as the total score is 12 (2+5+3+2), followed by C4 (20) and C3 (35). Next would be C2 (34+) and C6 (36+), this ranking being due to their unknown handshake capability (which had not been tried). C1(x) is the last to be tried because recent handshake attempts had failed for both "B" and "C." To call the three stations "A," "B," and "C," the sequence would be C5 (24), C3 (38), C2 (46+), C6 (62+), C4 (one x) (recently failed attempt), and finally C1 (two x).

If an additional selection factor is used, it will change the channel selection sequence.

NOTE: In the example, to call "D" and "E," the sequence would be C2, C3, C4, C5, C1, and C6. If a maximum limit of LQA < 14 is imposed on any path (to achieve a minimum circuit quality), only C2 and C3 would be initially selected for the linking attempt. Further, if the LQA limit was "lowered" to 10, C3 would be selected before C2 for the linking attempt.

If a broadcast to multiple stations is required, the TO section ("to" the station) scores are given priority.

NOTE: In the example, to broadcast to "B" and "C," the sequence would be C5(7), C4(9), C3(21), C2(7+), C6(12+), and C1(two x).

To select channels for listening for multiple stations, the FROM section ("from" the station) scores are given priority.

NOTE: In the example, to listen for "A" and "B," channel C2 (2) would be best, and if only four channels could be scanned, they should be C2, C3, C4, and C5.

A.5.4.7 Listen before transmit.

Before initiating a call or a sound on a channel, an ALE controller shall listen for a programmable time (T_wt) for other traffic, and shall not transmit on that channel if traffic is detected. Normally, a sound aborted due to detected traffic will be rescheduled, while for a call another channel shall be selected.

A.5.4.7.1 Listen-before-transmit duration.

The duration of the listen-before-transmit pause shall be programmable by the network manager. When the selected channel is known to be used only for ALE transmissions, the listen-before-transmit delay need be no longer than 2 T_rw. For other channels, at least 2 seconds shall be used. When an ALE controller was already listening on the channel selected for a transmission, the time spent listening on the channel may be included in the listen-before-transmit time.

A.5.4.7.2 Modulations to be detected.

The listen-before-transmit function shall detect traffic on a channel in accordance with A.4.2.2. This may be accomplished using any combination of internal signal detection and external devices that provide a channel busy signal to the ALE controller.

A.5.4.7.3 Listen before transmit override.

The operator shall be able to override both the listen-before-transmit pause and the transmit lockout (for emergency use).

A.5.5 Link establishment protocols.

An ALE controller shall control an attached HF SSB radio to support both manual and automatic link operation as described in the following paragraphs.

A.5.5.1 Manual operation.

The ALE controller shall support emergency control by the operator. Each ALE controller shall provide a manual control capability to permit an operator to directly operate the basic SSB radio in emergency situations. At all other times, the radio shall be under automated control, and the operator should operate the radio through its associated controller. The ALE controller's receiving and passive collection capability to be "always listening," such as monitoring for sounding signals or alerting the operator, shall not be impaired.

NOTE: This does not abrogate the manual push-to-talk operation required by 4.2.2.

A.5.5.2 ALE.

The fundamental protocol exchange for link establishment shall be the three-way handshake (see Appendix I for overview of Selective Calling). A three-way handshake is sufficient to establish a link between a calling station and a responding station. With the addition of slotted responses (described in A.5.5.4.2), the same call/response/acknowledgment sequence can also link a single calling station to multiple responding stations.

A.5.5.2.1 Timing.

The ALE system depends on a selection of timing functions for optimizing the efficiency and effectiveness of ALE. The primary timing functions and values as listed in table A-XV. Annex A defines the timing symbols and Annex B explains the timing analysis and computation.

A.5.5.2.2 ALE states.

An ALE controller may be referred to as being in one of three conceptual "states." See figure
A-28.

A.5.5.2.3 ALE channel selection.

A scanning calling station shall send ALE calls on its scanned channels in the order dictated by its channel selection algorithm. It shall link on the first channel it tries that supports a handshake with the called station(s).

A.5.5.2.3.1 Rejected channel.

If a channel is rejected after linking by the operator or controller as unsuitable, the ALE controller shall terminate the link in accordance with A.5.5.3.5 and shall update LQA data using measurements obtained during linking.

A.5.5.2.3.2 Busy channel.

During the scanning-calling cycle, a caller may encounter occupied channels and shall skip them to avoid interference to traffic and activity. After all available channels have been tried, if no contact has been successful, the caller should revisit the previously occupied channels and, if they are free, attempt to call.

A.5.5.2.3.3 Exhausted channel list.

If a calling station has exhausted all of its prearranged scan set channels and failed to establish a link, it shall immediately return to normal receive scanning (the available state). It shall also alert the operator (and networking controller if present) that the calling attempt was unsuccessful.

A.5.5.2.4 End of frame detection.

ALE controllers shall identify the end of a received ALE signal by the following methods. The controller shall search for a valid conclusion (TIS or TWAS, possibly followed by DATA and REP for a maximum of five words, or T_{x max}). The conclusion must maintain constant redundant word phase within itself (if a sound) and with associated previous words. The controller shall examine each successive redundant word phase (T_rw) following the TIS (or TWAS) for the first (of up to four) non-readable or invalid word(s). Failure to detect a proper word (or detection of an improper word) or detection of the last REP, plus the last word wait delay time, (T_lww or T_rw), shall indicate the end of the received transmission. The maximal acceptable terminator sequence is TIS (or TWAS), DATA, REP, DATA, REP.

A.5.5.3 One-to-one calling.

The protocol for establishing a link between two individual stations shall consist of three ALE frames: a call, a response, and an acknowledgment. The sequence of events, and the timeouts involved, are discussed in the following paragraphs using a calling station SAM and a called station JOE.

A.5.5.3.1 Sending an individual call.

After selecting a channel for calling, the calling station (SAM) shall begin the protocol by first listening on the channel to avoid "disturbing active channels," and then tuning. If the called station (JOE) is known to be listening on the chosen channel (not scanning), the calling station shall transmit a single-channel call that contains only a leading call and a conclusion (see upper frame in figure A-29). Otherwise, it shall send a longer calling cycle that precedes the leading call with a scanning call of sufficient length to capture the called station's receiver as it scans (lower frame in figure A-29). The duration of this scanning call shall be 2 T_rw for each channel that the called station is scanning. The scanning call section shall contain only the first word of the called station address, using a TO preamble, and repeated as necessary until the end of the scanning call section.

The entire called station address shall be used in the leading call section, and shall be sent twice (see figure A-29) using a TO preamble each time the first word is sent and DATA and REP as required for additional words.

Any message section CMDs shall be sent immediately following the leading call, followed by a conclusion containing the complete calling station address ("TIS SAM"). The calling station shall then wait a preset reply time to start to receive the called station's response. In the single-channel case, the wait for reply time shall be T_wr, which includes anticipated round trip propagation delay and the called station's turnaround time. In the multi-channel case, the calling station shall wait through a wait for reply and tune time (T_wrt), which also includes time for the called station to tune up on the chosen channel.

If the expected reply from the called station does not start to arrive within the preset wait for reply time (T_wr) or wait for reply and tune time (T_wrt), the linking attempt on this channel has failed. At this point, if other channels in the scan set have not been tried, the linking attempt will normally start over on a new channel. Otherwise, the ALE controller shall return to the available state, and the calling station's operator or networking controller shall be notified of the failed linking attempt.

A.5.5.3.2 Receiving an individual call.

When the called station (JOE) arrives on channel, sometime during its scan period T_s, and therefore during the calling station SAM's longer scan calling time T_sc, the called station shall attempt to detect ALE signaling within its dwell time. If ALE signaling is detected, and the controller achieves word sync, it shall examine the received word to determine the appropriate action.

If JOE reads "TO JOE" (or an acceptable equivalent according to protocols), the ALE controller shall stop scan, enter the linking state, and continue to read ALE words while waiting a preset, limited time T_wce for the calling cycle to end and the message or conclusion to begin.

• If the received word is potentially from a sound or some other protocol, the ALE controller shall process the word in accordance with that protocol.
• Otherwise, the ALE controller shall resume its previous state (e.g., available if it was scanning, linked if it was linked to another station).

While reading a call in the linking state, the called station shall evaluate each new received word. The controller shall immediately abort the handshake and return to its previous state upon the occurrence of any of the following:

• It does not receive the start of a quick-ID, message, or frame conclusion within T_wce, or the start of a conclusion within T_mmax after the start of the message section;
• Any invalid sequence of ALE word preambles is received, except that during receipt of a scanning call, up to three contiguous words containing uncorrectable errors shall be tolerated without causing rejection of the frame;
• The end of the conclusion is not detected within T_lww, (plus the additional multiples of T_rw if an extended address) after the first word of the conclusion.

If a quick-ID or a message section starts within T_wce, the called station, (JOE) shall attempt to read one or more complete messages within a new preset, limited time T_mmax

If a frame conclusion starts "TIS SAM," the called station shall wait and attempt to read the calling station's address (SAM) within a new preset, limited time T_xmax.

If an acceptable conclusion sequence with TIS is read, the called station shall start a "last word wait" timeout T_lww = T_rw while searching for additional address words (if any) and the end of the frame (absence of a detected word), which shall trigger its response. The called station will also expect the calling station to continue the handshake (with an acknowledgment) within the called station's reply window, T_wr, after its response. If TWAS is read instead, the called station shall not respond but shall return to its previous state immediately after reading the entire calling station address.

If all of the above criteria for responding are satisfied, the called station shall initiate an ALE response immediately after detecting the end of the call, unless otherwise directed by the operator or controller.

A.5.5.3.3 Response.

Upon receipt of a call that is addressed to one of its own self addresses (JOE), and which contains a valid calling station address in a TIS conclusion (SAM), the called station shall listen for other traffic on the channel. If the channel is not in use, the station shall tune up, send a response (figure A-30), and start its own reply timer T_wr. (The longer T_wrt timeout is not necessary unless the calling station will send its acknowledgment on a different channel than the one carrying the call, requiring re-tuning.) If the channel is in use, the ALE controller shall ignore the call and return to its previous state unless otherwise programmed.

If the calling station (SAM) successfully reads the beginning of an appropriate response ("TO SAM") starting within its timeout (either T_wr or T_wrt), it shall process the rest of the frame in accordance with the checks and timeouts described above for the call until it either aborts the handshake or receives the appropriate conclusion, which in this example is "TIS JOE."

Specifically, the calling station shall immediately abort the handshake upon the occurrence of any of the following:

• It does not receive an appropriate response calling cycle ("TO SAM") starting within the timeout;
• An invalid sequence of ALE word preambles occurs;
• It does not receive the appropriate conclusion ("TIS JOE") starting within T_lc (plus T_{m max}, if message included);
• The end of the conclusion is not detected within T_lww, (plus the additional multiples of T_rw if an extended address).

After aborting a handshake for any of the above reasons, the calling station will normally restart the calling protocol, usually on another channel.

If the calling station receives the proper conclusion from the called station ("TIS JOE") starting within T_lc (plus T_{m max}, if message included), it shall set a last word wait timeout as above and prepare to send an acknowledgment. If, instead, "TWAS JOE" is received, the called station has rejected the linking attempt, the calling station ALE controller shall abort the linking attempt and inform the operator of the rejected attempt.

A.5.5.3.4 Acknowledgment.

If all of the above criteria for an acceptable response are satisfied, and if not otherwise directed by the operator or networking controller, the calling station ALE controller shall alert its operator that a correct response has been received, send an ALE acknowledgment (see figure A-31), enter the linked state with the called station (JOE), and unmute the speaker.

A "wait for activity" timer T_wa shall be started (with a typical timeout of 30 seconds) that shall cause the link to be dropped if the link remains unused for extended periods (see A.5.5.3.5).

If the called station (JOE) successfully reads the beginning of an appropriate acknowledgment ("TO JOE") starting within its T_wr timeout, it shall process the rest of the frame in accordance with the checks and timeouts described above for the response until it either aborts the handshake or receives the appropriate conclusion, which in this example is "TIS SAM" or "TWAS SAM."

Specifically, the calling station shall immediately abort the handshake upon the occurrence of any of the following:

• It does not receive an appropriate response calling cycle ("TO JOE") starting within its T_wr timeout;
• An invalid sequence of ALE word preambles occurs;
• It does not receive the appropriate conclusion starting within T_lc after the start of the frame (plus T_{m max}, if message included);
• The end of the conclusion is not detected within T_lww, (plus the additional multiples of T_rw if an extended address).

If the handshake is aborted for any of the above reasons, the handshake has failed, and the called station ALE controller shall return to its pre-linking state. The called station shall notify the operator or controller of the failed linking attempt.

Otherwise, the called station shall enter the linked state with the calling station ("SAM"), alert the operator (and network controller if present), unmute the speaker, and set a wait-for-activity timeout T_wa.

NOTE 1: Although SAM's acknowledgment to JOE appears identical to a single-channel individual call from SAM to JOE, it does not cause JOE to provide another response to the acknowledgment (resulting in an endless "ping-pong" handshake) because SAM's acknowledgment arrives within a narrow time window (T_wr) after JOE's response, and an acknowledge (ACK) from SAM is expected within this window. If SAM's acknowledgment arrives late (after T_wr), however, then JOE must treat it as a new individual call (and shall therefore send a new response, if SAM concludes the frame with TIS).

NOTE 2: A typical one-to-one scanning call three-way handshake takes between 9 and 14 seconds.

A.5.5.3.5 Link termination.

Termination of a link after a successful linking handshake shall be accomplished by sending a frame concluded with TWAS to any linked station(s) which is (are) to be terminated. For example, "TO JOE, TO JOE, TWAS SAM" (when sent by SAM) shall terminate the link between stations SAM and JOE. JOE shall immediately mute and return to the available state, unless it still retains a link with any other stations on the channel. Likewise, SAM shall also immediately mute and return to the available state, unless it retains a link with any other stations on the channel.

A.5.5.3.5.1 Manual termination.

A means shall be provided for operators to manually reset a station, which shall mute the speaker(s), return the ALE controller to the available state, and send a link terminating (TWAS) transmission, as specified above, to all linked stations, unless this latter feature is overridden by the operator. (DO: provide a manual disconnect feature that drops individual links while leaving others in place.)

A.5.5.3.5.2 Automatic termination.

If no voice, data, or control traffic is sent or received by a station within a preset time limit for activity (T_wa), the ALE controller shall automatically mute the speaker, terminate the linked state with any linked stations, and return to the available state. The wait for the activity timer is mandatory, but shall also be capable of being disabled by the operator or network manager. This timed reset is not required to cause a termination (TWAS) transmission, as specified above. However, it is recommended that a termination be sent to reset the other linked stations(s) to immediately return them to the available state.

Termination during a handshake or protocol by the use of TWAS (or a timer) should cause the receiving (or timed-out) station to end the handshake or protocol, terminate the link with that station, re-mute, and immediately return to the available state unless it still retains a link with another station.

A.5.5.3.6 Collision detection.

While receiving an ALE signal, it is possible for the continuity of the received signal to be lost (due to such factors as interference or fading) as indicated by failure to detect a good ALE word at a T_rw boundary. When one or both Golay words of a received ALE word contain uncorrectable errors, the ALE controller shall attempt to regain word sync, with a bias in favor of words that arrive with the same word phase as the interrupted frame.

If word sync is reacquired but at a new word phase, this indicates that a collision has occurred. The interrupted frame shall be discarded, and the interrupting signal processed as a new ALE frame.

NOTE: Stations should be able to read interfering ALE signals, as they may contain useful (or critical) information, for which the station is "always listening."

A.5.5.4 One-to-many calling.

One station may simultaneously establish a multi-way link with multiple other stations using the protocols described in the following subparagraphs.

A.5.5.4.1 Slotted responses.

The simple three-way handshake used for individual links cannot be used for one-to-many calling because the responses from the called stations would collide with each other. Instead, a time-division multiple access (TDMA) scheme is used. Each responding station shall send its response in an assigned or computed time slot as described later for the particular one-to-many protocol.

At the end of a one-to-many call frame, the following events shall take place:

• The calling station shall set a wait-for-response-and-tune timeout (WRTT) that shall trigger its acknowledgment after the last response slot time has expired. The time allowed is denoted T_wrn. The value of T_wrn is described later for each one-to-many protocol.
• The called stations shall set their own WRTTs that bound their waiting times for an acknowledgment. To allow time for acquiring word sync during the leading call of the acknowledgment, the waiting time shall be set to T_wan = T_wrn + 2 T_rw.
• Each called station shall also set a slot wait timeout T_swt that shall trigger its response.
• The called stations shall tune as required during the slot immediately following the end of the call frame, called slot 0.

As each station's slot wait timer expires, it shall send its response and continue to await the expiration of its WRTT. Should that timer expire before the start of an acknowledgment from the calling station, the called station shall abort the linking attempt, and return to its pre-linking state.

A.5.5.4.1.1 Slotted response frames.

Slotted response frames shall be formatted identically to responses in the one-to-one calling protocol (see figure A-32), including a leading call, an optional message section, and a frame conclusion. A responding station shall conclude its response with TIS to accept the call, or TWAS to reject it. When the calling and responding addresses are one-word (as shown), slots are each 14 T_w, or about 1.8 seconds.

A.5.5.4.1.2 Slot widths.

Unless otherwise specified, all slots shall be 14 T_w in duration, which allows response frames with single-word addresses to propagate to and from the other side of the globe and use commonly available HF transceivers and tuners. When any slot is extended, all following slots shall be delayed commensurately.

• When the calling station address is longer than one word, every slot shall be extended by two T_rw (six T_w) per additional address word.
• When a called station address is longer than one word, its slot shall be extended by one T_rw (three T_w) per additional address word.
• Slots shall be extended by one T_rw (three T_w) for each ALE word to be sent in the message section of responses (including LQA CMD).

A.5.5.4.1.3 Slot wait time formula.

The general formula for determining the correct timing for slotted responses in nonminimum or nonuniform cases is as follows for a selected slot number denoted SN:

T_swt(SN) = SN x [5 T_w + 2 T_a (caller) + (optional message) T_m] + T_a (caller) +

m = SN-1

S T_a (m) (called)

m=1

Where T_a (caller) is the address length (an integer multiple of T_rw) of the calling station,

(optional message)T_m is an optional message section (same size for all slots), present if and only if requested in the call. T_a(m) (called) is the address length of the station that will respond in slot m. (Note that the length of slot 0 is determined by using the address length of the calling station.) The formula for the calling station wait for net reply timeout (T_wrn) is

T_wrn = T_swt (NS + 1)

where NS is the total number of slots; one is added to include slot zero.

The formula for the called station acknowledgment timer is

T_wan = T_wrn + 2 T_rw

A.5.5.4.1.4 Slotted response example.

The slotted response example is shown in figure A-33.

A.5.5.4.2 Star net calling protocol.

A net address is assigned to a set of net member stations, as described in A.5.2.4.4. The slot number and address to be used by each net member are preassigned and known to all net members.

A.5.5.4.2.1 Star net call.

A star net call is identical to a one-to-one call, except that the called station address is a net address, as shown in figure A-34. The calling station address shall be an individual station address (not a net or other collective address).

A.5.5.4.2.2 Star net response.

When an ALE controller receives a call that is addressed to a net address that appears in its self address memory (see A.4.3.2), it shall process the call using the same checks and timeouts as an individual call (see A.5.5.3.2). If the call is acceptable, it shall respond in accordance with A.5.5.4.1 using its assigned net member address and slot number for the net address that was called.

A.5.5.4.2.3 Star net acknowledgment.

A star net acknowledgment is identical to a one-to-one acknowledgment, except that the called station address is a net address.

An ALE controller that has responded to a net call shall process the acknowledgment from the calling station in accordance with A.5.5.3.4, except that the wait-for-response timeout value shall be the T_wan timeout from A.5.5.4.1.3. A TWAS acknowledgment from the calling station shall return the called ALE controller to its pre-linking state. If a TIS acknowledgment is received from the calling station, the called ALE controller shall enter the linked state with the calling station (SAM in this example), alert the operator (and network controller if present), unmute the speaker, and set a wait-for-activity timeout T_wa.

A.5.5.4.3 Star group calling protocol.

The group calling protocol extends the power of one-to-many calling to ad hoc collections of stations that have not been preprogrammed as a net. Nothing need be known about the stations except their individual addresses and scanned frequencies. Because a group is not set up in advance, stations must be able to derive group membership and slot parameters on the fly. Group membership is limited as follows:

• The total length of group member station addresses cannot exceed 12 ALE words.
• The set of unique first address words among group members cannot exceed five words.

A.5.5.4.3.1 Star group scanning call.

A group address is produced by combining individual addresses of the stations that are to form the group. During a scanning call, only the first word(s) of addresses shall be sent, just as for individual or net calls. The set of unique first address words for the group members shall be sent repeatedly in rotation until the end of T_sc. These address words shall alternate between THRU and REP preambles (see figure A-35 for a sample group consisting of BOB, EDGAR, and SAM).

When group member addresses share a common first word, that word shall be sent only once during T_sc. A limit of five unique first words may be sent in rotation during T_sc.

A.5.5.4.3.2 Star group leading call.

During T_lc, the complete addresses of the prospective group members shall be sent, using TO preambles as usual. Up to 12 address words total are allowed for the full addresses of group members, so T_lc in a group call may last up to 24 T_rw. Note in figure A-34 that when a TO word would follow another TO word, a REP preamble must be used, but when a TO follows any other word it shall remain a TO.

A.5.5.4.3.3 Star group call conclusion.

The optional message section and the conclusion of a star group call shall be in accordance with A.5.2.5.

A.5.5.4.3.4 Receiving a star group call.

Slots shall be derived for group call responses by noting the order in which individual addresses appear in the call.

a. When an ALE controller pauses on a channel carrying a group scanning call, it will read either a THRU or a REP preamble. If the address word in this first received word matches the first word of one of its individual addresses, the ALE controller shall stay to read the leading call. Otherwise, it shall continue to read first address words until it finds:

• a match with the first word of a self address, or
• a repetition of a word it has already seen, or
• five unique words.

(In the latter two cases, the station is not being called and the ALE controller shall return to the available or linked state as appropriate.)

b. When T_lc starts, an ALE controller potentially addressed in the scanning call shall watch for its complete address. If found, a slot counter shall be set to 1 and incremented for each address that follows it. If that address is found again (as it should be, because the address list is repeated in T_lc), the counter shall be then reset to 1, and incremented for each following address as before. The number of words in each following address shall also be noted for use in computing T_swt.

c. The message section (if any) and the frame conclusion shall processed in accordance with A.5.5.3.2.

In the event that an addressed ALE controller arrives on channel too late to identify the size of the called group, it will be unable to compute the correct T_wan. In this situation, it shall use a default value for T_wan, which is equal to the longest possible group call of twelve one-word addresses. It will, however, have computed its correct slot number because to have received its own address it must also have received the addresses that followed that self address in the leading call.

A.5.5.4.3.5 Star group slotted responses.

Slotted responses shall be sent and checked in accordance with A.5.5.4.1, using the derived slot numbers and the self address contained in the leading call.

A.5.5.4.3.6 Star group acknowledgment.

The acknowledgment in a group call handshake shall be addressed to any subset of the members originally called, and is usually limited to those whose responses were heard by the calling station. The leading call of the acknowledgment shall include the full addresses of the stations addressed, sent twice, using the same syntax as in the call (A.5.5.4.3.2).

An ALE controller that responded to a group call shall await acknowledgment and process an incoming acknowledgment in accordance with A.5.5.3.4, with the following exceptions:

• The wait-for-response timeout value shall be the T_wan timeout from A.5.5.4.1.3, not T_wr.
• Self address detection shall search through the entire leading call group address.

An ALE controller that responded but was not named in the acknowledgment shall return to its pre-linking state. An ALE controller that is addressed in the acknowledgment shall proceed as follows:

• A TWAS acknowledgment from the calling station shall return the called ALE controller to its pre-linking state.
• If a TIS acknowledgment is received from the calling station, the called ALE controller shall enter the linked state with the calling station (SAM in this example), alert the operator (and network controller if present), unmute the speaker, and set a wait-for-activity timeout T_wa.

A.5.5.4.3.7 Star group call example.

In the example group call in figure A-35, SAMUEL will respond in slot 1, with T_swt = 14 T_w (the one-word address JOE causes slot 0 to be 14 T_w). EDGAR will respond in slot 2, with T_swt = 14 + 17 T_w = 31 T_w (slot 1 is 17 T_w because of SAMUEL's two-word address). BOB will respond in slot 3, with T_swt = 48 T_w. JOE will send an acknowledgment after 62 T_w.

A.5.5.4.3.8 Multiple self addresses in group call.

If a station is addressed multiple times in a group call, even by different addresses, it shall properly respond to at least one address.

NOTE: The fact that the called station has multiple addresses may not be known to the caller. In some cases, it would be confusing or inappropriate to respond to one but not another address. Redundant calling address conflicts can be resolved after successful linking, if there is a problem.

A.5.5.4.4 Allcall protocol.

An AllCall requests all stations hearing it to stop and listen, but not respond. The AllCall special address structure(s) (see A.5.2.4.7) shall be the exclusive member(s) of the scanning call and the leading call, and shall not be used in any other address field or any other part of the handshake. The global AllCall address shall appear only in TO words. Selective AllCalls with more than one selective AllCall address, however, shall be sent using group addressing, using THRU during the scanning call and TO during the leading call.

An AllCall pertains to an ALE controller when it is a global AllCall, or when a selective AllCall specifies a character that matches the last character of any self address assigned to that station. Upon receipt of a pertinent AllCall, an ALE controller shall temporarily stop scanning and listen for a preset limited time, T_{cc max}.

• If a message section or frame conclusion does not arrive within T_{cc max}, the controller shall automatically resume scanning.
• If a quick-ID (an address beginning with a FROM word immediately after the calling cycle) arrives, the pause for the message section shall be extended for no more than five words (5 T_rw), and if a CMD does not arrive, the controller shall resume scanning.
• If a message arrives (indicated by receipt of a CMD), the controller shall pause for a preset limited time, T_{m max} to read the message. If the frame conclusion does not arrive within T_{m max}, the controller shall automatically resume scanning. If a conclusion arrives (indicated by receipt of a TIS or TWAS), the controller shall pause (for a preset limited time, T_{x max}) to read the caller's address. If the end of the signal does not arrive within T_{x max}, the controller shall automatically resume scanning.

If a pertinent AllCall frame is successfully received and is concluded with a TIS, the controller shall enter the linked state, alert the operator, unmute its speaker and start a wait-for-activity timeout. If an AllCall is successfully received with a TWAS conclusion, the called controller shall automatically resume scanning and not respond (unless otherwise directed by the operator or controller).

If a station receiving an AllCall desires to attempt to link with the calling station, the operator may initiate a handshake within the pause after a TIS conclusion. Note that in all handshakes (the initial AllCall does not constitute a handshake), the AllCall address shall not be used. To minimize possible adverse effects resulting from overuse or abuse of AllCalls, controllers shall have the capability to ignore AllCalls. Normally AllCall processing should be enabled.

A.5.5.4.5 AnyCall protocol.

An AnyCall is similar to an AllCall, but it instead requests responses. Use of the AnyCall special address structures is identical to that for the AllCall special address structures. Upon receipt of a pertinent AnyCall, an ALE controller shall temporarily stop scanning and examine the call identically to the procedure for AllCalls, including the T_{cc max}, T_{m max}, and T_{x max} limits.

If the AnyCall is successfully received, and is concluded with TIS, the controller shall enter the linking state and automatically generate a slotted response in accordance with A.5.5.4.1 and the following special procedure:

• Because neither preprogrammed nor derived slot data are available, the controller shall randomly select a slot number, 1 through 16.
• Each slot shall be 20 T_w (2613.33...ms) wide, unless the calling station requests LQA responses, in which case the slots shall expand by 3 T_w to 23 T_w to accommodate the CMD LQA message section.
• The controller shall compute values for T_swt and T_wan using this slot width and its random slot number.
• Slot 0 shall be used for tuning, as usual for slotted response protocols.
• Upon expiration of its T_swt timeout, the controller shall send a standard star net response consisting of TO (with the address of the caller) and TIS (with the address of the responder), with the LQA CMD included if requested. Responders shall use a self address no longer than five words minus twice the caller address length. (For example, if the caller address is two words, the responder shall use a one-word address.) The AnyCall special address shall not be sent.

In this protocol, collisions are expected and tolerated. The station sending the AnyCall shall attempt to read the best response in each slot.

Upon receipt of the slotted responses, the calling station shall transmit an ACK to any subset of stations whose responses were read, using an individual or group address. The AnyCall special address shall not be used in the acknowledgment. The caller selects the conclusion of its ACK to either maintain the link for additional interoperation and traffic with the responders (TIS), or return everyone to scan (TWAS), as appropriate to the caller's original purpose.

An ALE controller that responded to an AnyCall shall await and process the acknowledgment in accordance with A.5.5.4.3.6.

To minimize possible adverse effects resulting from overuse or abuse of AnyCalls, controllers shall have the capability to ignore AnyCalls. Normally AnyCall processing should be enabled.

A.5.5.4.6 Wildcard calling protocol.

Wildcard addresses shall be the exclusive members of a calling cycle in a call, and shall not be used in any other address sequence in the ALE frame or handshake. The span (number of cases possible) of the wildcard(s) used should be minimized to only the essential needs of the user(s).

Calls to wildcard addresses that conclude with TWAS shall be processed identically to the AllCall protocol.

Responses to wildcard calls that conclude with TIS shall be sent in pseudorandomly-selected slots in accordance with the AnyCall protocol.

As in both the AllCall and AnyCall, the controller shall be programmable to ignore wildcard calls, but wildcard call processing should normally be enabled.

A.5.6. ALE control functions (CMDs other than AMD, DTM, and DBM).

In addition to automatically establishing links, stations shall have the capability to transfer information within the orderwire, or message, section of the frame. This section describes these messages, including data, control, error checking, networking, and special purpose functions. Table A-XVI provides a summary of the CMD functions.

NOTE: For critical orderwire messages that require increased protection from interference and noise, several ALE techniques are available. Any message may be specially encoded off-line and then transmitted using the full 128 ASCII CMD data DTM mode (which also accepts random data bits). Larger blocks of information may be Golay FEC coded and deeply interleaved using the CMD DBM mode. Both modes have an automatic repeat request (ARQ) error-control capability. Integrity of the data may be ensured using the CMD cyclic redundancy check (CRC) mode (see A.5.6.1). In addition, once a link has been established, totally separate equipment, such as heavily coded and robust modems, may be switched onto the rf link in the normal circuit (traffic-bearing) mode.

A.5.6.1 CRC.

This special error-checking function is available to provide data integrity assurance for any form of message in an ALE call.

NOTE: The CRC function is optional, but mandatory when used with the DTM or DBM modes.

The 16-bit frame check sequence (FCS) and method as specified by FED-STD 1003 shall be used herein. The FCS provides a probability of undetected error of 2^-16, independent of the number of bits checked. The generator polynomial is

X¹⁶ + X¹² + X⁵ + 1

and the sixteen FCS bits are designated

(MSB) X¹⁵, X¹⁴, X¹³, X¹²...X¹, X⁰ (LSB)

The ALE CRC is employed two ways: within the DTM data words, and following the DBM data field, described in paragraphs A.5.7.3 and A.5.7.4, respectively. The first, and the standard, usages are described in this section.

The CMD CRC word shall be constructed as shown in table A-XVII. The preamble shall be CMD (110) in bits P3 through P1 (W1 through W3). The first character shall be "x" (1111000), "y" (1111001), "z" (1111010), or "{" (1111011) in bits C1-7 through C1-1 (W4 through W10). Note that four identifying characters result from FCS bits X¹⁵ and X¹⁴ which occupy C1-2 and C1-1 (W9 and W10) in the first character field respectively. The conversion of FCS bits to and from ALE CRC format bits shall be as described in table A-XVII where X¹⁵ through X⁰ correspond to W9 through W24.

The CMD CRC message should normally appear at the end of the message section of a transmission, but it may be inserted within the message section (but not within the message being checked) any number of times for any number of separately checked messages, and at any point except the first word (except as noted below). The CRC analysis shall be performed on all ALE words in the message section that precede the CMD CRC word bearing the FCS information, and which are bounded by the end of the calling cycle, or the previous CMD CRC word, whichever is closest. The selected ALE words shall be analyzed in their non-redundant and unencoded (or FEC decoded) basic ALE word (24-bit) form in the bit sequence (MSB) W1, W2, W3, W4...W24 (LSB), followed by the unencoded bits W1 through W24 from the next word sent (or received), followed by the bits of the next word, until the first CMD CRC is inserted (or found). Therefore, each CMD CRC inserted and sent in the message section ensures the data integrity of all the bits in the previous checked ALE words, including their preambles. If it is necessary to check the ALE words in the calling cycle (TO) preceding the message section, an optional calling cycle CMD CRC shall be used as the calling cycle terminator (first FROM or CMD), shall therefore appear first in the message section, and shall analyze the calling cycle words in their simplest (T_c), nonredundant and nonrotated form. If it is necessary to check the words in a conclusion (TIS or TWAS), an optional conclusion CRC shall directly precede the conclusion portion of the call, shall be at the end of the message section, and shall itself be directly preceded by a separate CMD CRC (which may be used to check the message section or calling cycle, as described herein). Stations shall perform CRC analysis on all received ALE transmissions and shall be prepared to compare analytical FCS values with any CMD CRC words which may be received. If a CRC FCS comparison fails, an ARC (or operator initiated) or other appropriate procedure may be used to correct the message.

A.5.6.2 Power control (optional).

The power control orderwire function is used to advise parties to a link that they should raise or lower their rf power for optimum system performance. The power control CMD word format shall be as shown in figure A-36. The KP control bits shall be used as shown in table XVIII.

3	7	3	6	5
CMD	1110000 ('p': power control)	KP1-3	Power	(reserved)

The procedure shall be:

a. When KP₃ is set to 1, the power control command is a request to adjust the power from the transmitter. If KP₂ is 1, the adjustment is relative to the current operating power, i.e., to raise (KP₁ = 1) or lower (KP₁ = 0) power by the number of dB indicated in the relative power field. If KP₂ is 0, the requested power is specified as an absolute power in dBW.

b. When KP₃ is set to 0, the power control command reports the current power output of the transmitter, in dB relative to nominal power if KP₂ is 1, or in absolute dBW if KP₂ is 0.

c. KP₁ shall be set to 0 whenever KP₂ is 0.

d. Normally, a station receiving a power control request (KP₃ = 1) should approximate the requested effect as closely as possible, and respond with a power report (KP₃ = 0) indicating the result of its power adjustment.

A.5.6.3 Channel related functions.

The channel related functions are defined in the following subparagraphs.

A.5.6.3.1 Channel designation.

When two or more stations need to explicitly refer to channels or frequencies other than the one(s) in use for a link, the following encodings shall be used. A frequency is designated using binary-coded-decimal (BCD). The standard frequency designator is a five-digit string (20 bits), in which the first digit is the 10 megahertz (MHz) digit, followed by 1 MHz, 100 kilohertz (kHz), 10 kHz, and 1 kHz digits. A frequency designator is normally used to indicate an absolute frequency. When a bit in the command associated with a frequency designator indicates that a frequency offset is specified instead, the command shall also contain a bit to select either a positive or a negative frequency offset.

A.5.6.3.2 Frequency designation.

A channel differs from a frequency in that a channel is a logical entity that implies not only a frequency (or two frequencies for a full-duplex channel), but also various operating mode characteristics, as defined in A.4.3.1. As in the case of frequency designators, channels may be specified either absolutely or relatively. In either case, a 7-bit binary integer shall be used that is interpreted as an unsigned integer in the range 0 through 127. Bits in the associated command shall indicate whether the channel designator represents an absolute channel number, a positive offset, or a negative offset.

a. The frequency select CMD word shall be formatted as shown in figure A-37. A frequency designator (in accordance with A.5.6.3.1) is sent in a DATA word immediately following the frequency select CMD; bit W4 of this DATA word shall be set to 0, as shown.

3	7	6	4	4
CMD	1100110 ('f': frequency)	Control	100 Hz	10 Hz

3	1	4	4	4		4	4
DATA	0	Frequency Designator
		10 MHz	1 MHz		100 kHz	10 kHz		1 kHz

b. The 100 Hz and 10 Hz fields in the frequency select CMD word contain BCD digits that extend the precision of the standard frequency designator. These digits shall be set to 0 except when it is necessary to specify a frequency that is not an even multiple of 1 kHz (e.g., when many narrowband modem channels are allocated within a 3 kHz voice channel).

c. The control field shall be set to 000000 to specify a frequency absolutely, to 100000 to specify a positive offset, or to 110000 to specify a negative offset.

d. A station receiving a frequency select CMD word shall make whatever response is required by an active protocol on the indicated frequency.

A.5.6.3.3 Full-duplex independent link establishment (optional).

Full duplex independent link establishment is an optional feature; however, if this option is selected the transmit and receive frequencies for use on a link shall be negotiated independently as follows:

a. The caller shall select a frequency believed to be propagating to the distant station (the prospective responder) and places a call on that frequency. The caller embeds a frequency select CMD word in the call to ask the responder to respond on a frequency chosen for good responder-to-caller propagation (probably from sounding data in the caller's LQA matrix).

b. If the responder hears the call, it shall respond on the second frequency, asking the caller to switch to a better caller-to-responder frequency by embedding a frequency select CMD word in its response (also based upon sounding data).

c. The caller shall send an acknowledgment on the frequency chosen by the responder (the original frequency by default), and the full duplex independent link is established.

A.5.6.3.4 LQA polling (optional).

See MIL-STD-187-721.

A.5.6.3.5 LQA reporting (optional).

See MIL-STD-187-721.

A.5.6.3.6 LQA scan with linking (optional).

See MIL-STD-187-721.

A.5.6.3.7 Advanced LQA (optional).

See MIL-STD-187-721.

A.5.6.4 Time-related functions.

A.5.6.4.1 Tune and wait.

The CMD tune and wait special control function directs the receiving station(s) to perform the initial parts of the handshake, up through tune-up, and wait on channel for further instructions during the specified time limit. The time limit timer is essentially the WRTT as used in net slotted responses where its value T_wrn is set by the timing information in the special control instruction, and it starts from the detected end of the call. The CMD tune and wait instruction shall suppress any normal or preset responses. Except for the tune-up itself, the receiving station(s) shall make no additional emissions, and they shall quit the channel and resume scan if no further instructions are received.

NOTE: This special control function enables very slow tuning stations, or stations that must wait for manual operator interaction, to effectively interface with automated networks.

The CMD tune and wait shall be constructed as follows and as shown in table A-XIX. The preamble shall be CMD (110) in bits P3 through P1 (W1 through W3). The first character (C1) shall be "t" (1110100) in bits C1-7 through C1-1 (W4 through W10) and "t" (1110100) in bits C2-7 through C2-1 (W11 through W17), for "time, tune-up." The "T" time bits TB7 through TB1 (W18 through W24) shall be values selected from table A-XX, and limited as shown in table A-XXI. The lowest value (00000) shall cause the tuning to be performed immediately, with zero waiting time, resulting in immediate return to normal scan after tuning.

A.5.6.4.2 Scheduling commands.

These special control functions permit the manipulation of timing in the ALE system. They are based on the standard "T" time values, presented in table A-XX, which have the following ranges based on exact multiples of T_w (130.66...ms) or T_rw (392 ms).

• 0 to 4 seconds in 1/8 second (T_w) increments
• 0 to 36 seconds in 1 second (3 T_rw) increments
• 0 to 31 minutes in 1 minute (153 T_rw) increments
• 0 to 29 hours in 1 hour (9184 T_rw) increments

There are several specific functions that utilize these special timing controls. All shall use the CMD (110) preamble in bits P3 through P1 (W1 through W3). The first character is "t" (1110100) for "time." The second character indicates the function as shown in table A-XXI. The basic structure is the same as in table A-XIX.