TIA EIA IS-823-A-2001 TTY TDD Extension to TIA EIA 136-410 Enhanced Full Rate Speech Codec (Revision of TIA EIA IS-823)《TIA EIA 136-410全速率语音编码解码器TTY TDD分机[代替 TIA TIA EIA IS-823]》.pdf

《TIA EIA IS-823-A-2001 TTY TDD Extension to TIA EIA 136-410 Enhanced Full Rate Speech Codec (Revision of TIA EIA IS-823)《TIA EIA 136-410全速率语音编码解码器TTY TDD分机[代替 TIA TIA EIA IS-823]》.pdf》由会员分享，可在线阅读，更多相关《TIA EIA IS-823-A-2001 TTY TDD Extension to TIA EIA 136-410 Enhanced Full Rate Speech Codec (Revision of TIA EIA IS-823)《TIA EIA 136-410全速率语音编码解码器TTY TDD分机[代替 TIA TIA EIA IS-823]》.pdf（28页珍藏版）》请在麦多课文档分享上搜索。

1、TIA/EIA INTERIM STANDARD TTY/TDD Extension to TIA/EIA 136-410 Enhanced Full Rate Speech Codee TIA/EIA/IS-823-A (Revision of TIAIEIAIIS-823) SEPTEMBER 2001 TELECOMMUNICATIONS INDUSTRY ASSOCIATION +. . I , Repiesenting the telecommunications industiy in association with the Electionic Industiies fianc

2、e Elsctrrrnk Industries Alliance COPYRIGHT EIA Telecommunications Industry AssociationLicensed by Information Handling ServicesNOTICE TIA/EIA Engineering Standards and Publications are designed to serve the public interest through eliminating misunderstandings between manufacturers and purchasers, f

3、acilitating interchangeability and improvement of products, and assisting the purchaser in selecting and obtaining with minimum delay the proper product for his particular need. Existence of such Standards and Publications shall not in any respect preclude any member or nonmember of TIA/EIA from man

4、ufacturing or selling products not conforming to such Standards and Publications, nor shall the existence of such Standards and Publications preclude their voluntary use by those other than TIA/EIA members, whether the standard is to be used either domestically or internationally. Standards and Publ

5、ications are adopted by TIA/EIA in accordance with the American National Standards Institute (ANSI) patent policy. By such action, TIA/EIA does not assume any liability to any patent owner, nor does it assume any obligation whatever to parties adopting the Standard or Publication. TIA/EIA INTERIM ST

6、ANDARDS TIA/EIA Interim Standards contain information deemed to be of technical value to the industry, and are published at the request of the originating Committee without necessarily following the rigorous public review and resolution of comments which is a procedural part of the development of a

7、TIA/EIA Standard. TIA/EIA Interim Standards should be reviewed on an annual basis by the formulating Committee and a decision made on whether to proceed to develop a TIA/EIA Standard on this subject. TIA/EIA Interim Standards must be cancelled by the Committee and removed from the TIA/EIA Standards

8、Catalog before the end of their third year of existence. Publication of this TIA/EIA Interim Standard for trial use and comment has been approved by the Telecommunications Industry Association. Distribution of this TIA/EIA Interim Standard for comment shall not continue beyond 36 months from the dat

9、e of publication. It is expected that following this 36 month period, this TIA/EIA Interim Standard, revised as necessary, will be submitted to the American National Standards Institute for approval as an American National Standard. Suggestions for revision should be directed to: Standards that a ce

10、rtain course of action is preferred but not necessarily required; or that (in the negative form) a certain possibility or course of action is discouraged but not prohibited. “May” and “need not” indicate a course of action permissible within the limits of the standard. “Can” and “cannot” are used fo

11、r statements of possibility and capability, whether material, physical, or causal. Overview The following provides a method for reliably transporting the 45.45 bps and 50 bps Baudot code in TIMEIA-136-410, making digital wireless telephony accessible to TTY/TDD users. The following extension is robu

12、st to frame and bit errors and is completely interoperable with the pre-existing TIMEIA- 136-41 O standard and its annexes. The solution supports voice carryover/hearing carryover (VCO/HCO). VCO I allows a TTY/TDD user to switch between receiving TTY and talking into the phone. Similarly, HCO allows

13、 a user to switch between transmitting TTY characters and picking up the phone to listen. When Baudot tones are not present, the vocoder operates as usual; there is no modification or added delay to the voice path when speech is present. The TTY/TDD audio solution transports Baudot signals through t

14、he vocoder by detecting the characters that are being transmitted by the TTY/TDD in the encoder and conveying those characters to the decoder. Because one Baudot character spans at least 7 speech processing frames, the character being transmitted shall be sent a minimum of 6 times to the decoder, al

15、lowing the decoder to reliably regenerate the character despite frame errors and random bit errors in the speech packet. I Page 5 COPYRIGHT EIA Telecommunications Industry AssociationLicensed by Information Handling ServicesTIMEIAAS- 823 -A The TTY characters are concealed in the speech packet in a

16、way that interoperates with legacy vocoders that have not been modified. This is made possible because, when Baudot tones are present, the TTY information replaces some of the pitch lag bits for the adaptive codebook (ACB) and the ACB gain is restricted to a few low values so that an unmodified deco

17、der can decode audio information from the received packet. Even though the packet has been modified, there is enough information in the speech packet for an unmodified decoder to reconstruct the Baudot tones at least as well as if the encoder was not modified. A decoder modified with this extension

18、maintains a history buffer to monitor the ACB gain and pitch lag in the speech packets. When the decoder detects that the ACB gain has been forced low and the pitch lag contains information consistent with TTY, the decoder stops decoding speech and begins regenerating the Baudot tones. When the deco

19、der stops detecting TTY information, it resumes processing speech. When Baudot tones are not detected, the modified vocoder operates on speech exactly the same way as the unmodified vocoder. Furthermore, the TTY processing does not add any additional delay to the speech path. I 2 3 4 5 6 7 8 9 10 II

20、 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 TTY/T D D Extension 3.1 The TTY processing in the encoder shall process the RX PCM one frame at a time and label each frame as either NON-TTY, TTY-SILENCE, or as a TTY character. The vocoder will be in one of two st

21、ates, TTY-MODE or NON-TTY-MODE. In the absence of Baudot tones, the encoder and decoder shall be in the NON-TTY-MODE and the encoder and decoder shall process the frame as speech. When Baudot tones are present, the encoder and decoder shall enter TTYMODE and process the TTY information as described

22、below. It is recommended that a mechanism exist to disable the TTY/TDD extension in the vocoder, reverting the vocoder to its unmodified state. TTY Onset Procedure The TTY Onset Procedure describes the process by which the vocoder shall transition from speech mode to TTY mode. 3.1 .I Encoder TTY Ons

23、et Procedure The enccder shall send the TT Y SILENCE niessage to tile deroder within detecting the first 3 bits of rhe Baudot code, or sooner, For the very first onset of Baudot tones during a call, 1.e for tile fk Baudot ciiaracter oniy, the encoder may delay sending its TTY-SILENCE message until i

24、t detects the transition from a space tone f 1 SUO HZ) to 3 marl THRESH. If that threshold is exceeded, the dit is labeled as either a mark or a space, which ever one has the most energy. If a mark or space is not detected, the total energy is compared to a silence threshold. If the total energy is

25、below this threshold, the dit is labeled as TTY-SILENCE. If none of the thresholds are met, the dit is labeled as NON-TTY. 4.1.4 di t-to-asci i 0 The routine dit-to-ascii() takes the 10 dits generated by baudot-todito each frame and returns the state of the TTY encode processing and the character in

26、formation as it is detected. The routine get-tty-state() transitions the encoder to NON-TTYMODE or TTY-ONSET based on the detected TTY bits. TTY-MODE is entered when the TTY bits form a character. The dits are first grouped into TTY bits by get-tty_bit(). Page 12 I COPYRIGHT EIA Telecommunications I

27、ndustry AssociationLicensed by Information Handling ServicesTIMEIAAS- 823 -A Once a character is found, the character information and its header are sent to the decoder a minimum of 6 frames and a maximum of 21 frames. The constant FRAMING-HANGOVER dictates the maximum number of times the informatio

28、n for the same character is sent. If a new character is framed before FRAMING-HANGOVER is reached, the information for the old character is terminated and the new information is sent to the decoder. O Memory Bit I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 3

29、3 34 35 36 37 38 39 40 41 42 43 44 1 2 3 4 5 6 7 8 Start LSB MSB Stop Next Bit Data O Data 1 Data2 Data3 Data4 Bit Bit 4.1.5 get-tty-bi t() The routine get-tty-bit() forms TTY bit detections from the dit detections. Dits are classified as either LOGIC-O, LOGIC-1, or UNKNOWN. A nominal bit consists o

30、f 11 dits for 45.45 baud and 10 dits for 50 baud. A bit is not required to have a continuous run of LOGIC-Os or LOGIC-1s in order to be detected. Spurious UNKNOWN detections are permitted, up to a threshold. A TTY bit is searched by looking at the dits within a variable length sliding window. The wi

31、ndow varies in length from 8 dits to 13 dits. The number of LOGIC-Os, LOGIC-ls, and UNKNOWN dit detections are counted in the window. The dits in the search window must meet the following criteria in order to detect a TTY bit: Heuristics are also applied so that the sliding window is centered over t

32、he TTY bit being detected. If all of the thresholds are met, a “O” (“1”) TTY bit is declared, and the overall length of the bit and the number of bad dits within the window are recorded. If the dits in the search window do not meet the criteria for a TTY bit, a gap between TTY bits is declared and t

33、he gap is recorded. The window is slid by one dit and the search is repeated. The length of the search window is allowed to vary. The length is adjusted depending on the previous characters baud rate, and where the markhpace transitions occur. Two history buffers are maintained to record the detecte

34、d TTY bits and gaps. The array tq-bit-hist maintains a history of the bits that are detected and tty-bit-len-hist stores the lengths, in dits, of the gaps and TTY bits that are detected. Both arrays are 9 elements long, there is one element for the start bit, five for the data bits, one for the stop

35、 bit, and one for the memory bit, the bit before the start bit, as depicted in Figure 3. The last element in the array is used to record partially detected TTY bits. Figure 3: TTY Bit History Buffer Minimum of 6 LOGIC-O (LOGIC-1) dits Maximum of 2 LOGIC-1 (LOGIC-O) dits Maximum of 5 UNKNOWN dits Whe

36、n a TTY bit or a gap is detected, its value is recorded in tty-bit-hist and its length is stored in tq-bit-len-hist 1. The length is used by get-tty-char() to threshold the length of a candidate character, and by tq-rate() to determine if a detected character is 45.45 baud or 50 baud. Since the leng

37、th of the memory bit is irrelevant, tq-bit-len-histO is used to count the number of NON-TTY and TTY-SILENCE dits that were detected within the TTY bits. Because the speech frames may not coincide with the boundaries of the TTY bits, it is possible that a bit may straddle two speech frames. It is pos

38、sible, therefore, that the sliding window may contain only a partial bit within a frame. These partial detections are recorded in element 8 of the TTY bit history buffers, labeled “Next Bit” in Figure 3. Page 13 I COPYRIGHT EIA Telecommunications Industry AssociationLicensed by Information Handling

39、ServicesTIMEIAAS- 823 -A 4.1.6 get-tty-char() The routine get-tq-char() checks if the detected bits form a TTY character. following conditions must be met in order for a character to be declared. 0 0 0 0 0 If all of the conditions are met. a character is declared. The The bit preceding the start bit

40、 must not be a “O”. The start bit must be a “O” The stop bit must be a “1” The length of the candidate character, in dits, must be within a maximum and minimum threshold. The number of bad dit detections within the character must be within a threshold. I 2 3 4 5 6 I 8 9 10 II 12 13 14 15 16 11 18 19

41、 20 21 22 23 24 25 26 21 28 29 30 31 32 33 34 35 36 31 38 39 40 41 42 43 44 45 46 41 4.1.7 tty-rate() After a character has been detected, tty-rate() determines the baud rate of the character by counting the length, in dits, of the start bit and the five data bits. The length of the stop bit is not

42、used because its length is too variable. A character with a length of 63 dits or greater is declared 45.45 baud, otherwise it is declared 50 baud. A hangover of three characters is maintained before tty-rate() will switch from one baud rate to the other. That is to say, if the baud rate changes in t

43、he middle of a call, three consecutive characters must be detected at the new baud rate in order for tq-rate() to declare the new baud rate. At call startup time, the baud rate is initialized to 45.45 baud, the most common baud rate in North America. If a 50 baud call is initiated, it will take at l

44、east 3 characters before the baud rate will switch to 50 baud, because of the hangover. 4.1.8 get-tty-state() This routine is responsible for changing the state of the TTY encoder processing. There are 3 states, NON-TTYMODE, TTY-ONSET, and TTYMODE. Get-tq-stateO is responsible for determining NON-TT

45、YMODE and TTY-ONSET. In order to prevent false alms when Baudot tones are not present, a tougher rule is applied to declare TTY-ONSET for the very first time in a call. After the first character is declared, a more relaxed rule is allowed. The first rule requires that a “O” bit is followed by a “1”

46、bit. This rule is considered the most stringent of the two rules because it requires the presence of both the space tone and the mark tone and for the tones to be the correct duration. This test must be met in order to declare TTY-ONSET for the first time. After the first TTY character is detected,

47、TTY-ONSET may be declared either by the transition from a “O” to a “l”, or when three TTY bits are detected in a row. Although the fiist rule is less likely to false alm, it is susceptible to creating added characters because TTY-ONSET may not get declared in time to alert the decoder to mute before

48、 the Baudot tones pass through the speech path. The second rule is more prone to false alms, but alerts the decoder to mute in time to prevent double characters. By requiring the first rule to be satisfied for the first TTY character in the call, a false alarm is unlikely to occur in non-TTY calls a

49、nd the second rule is used only after the call has been identified as a TTY call. NON-TTY-MODE is declared by get-tty-state() whenever a non-TTY bit is detected or when a “O” bit occupies the bit preceding the start bit. 4.2 TTY/TDD Decoder Processina The TTY/TDD decoder subroutines and their hierarchy are shown in Figure 4. The initialization function init-tty-dec() initializes all of the state variables and static arrays for the TTY decoder processing. Fiaure 4: TTY Decoder Subroutines 48 Page 14 COPYRIGHT EIA Telecommunications Industry AssociationLicensed by Information H