1、 TIA DOCUMENT Minimum Performance Specification for Text Telephone Signal Detector and Text Telephone Signal Regenerator TIA-889-A (Revision of TIA/EIA/IS-889) December 2003 TELECOMMUNICATIONS INDUSTRY ASSOCIATION The Telecommunications Industry Association represents the communications sector of Co
2、pyright Electronic Industries Alliance Provided by IHS under license with EIANot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NOTICE TIA Engineering Standards and Publications are designed to serve the public interest through eliminating misunderstandings between ma
3、nufacturers and purchasers, facilitating interchangeability and improvement of products, and assisting the purchaser in selecting and obtaining with minimum delay the proper product for their particular need. The existence of such Publications shall not in any respect preclude any member or non-memb
4、er of TIA from manufacturing or selling products not conforming to such Publications. Neither shall the existence of such Documents preclude their voluntary use by non-TIA members, either domestically or internationally. TIA DOCUMENTS TIA Documents contain information deemed to be of technical value
5、 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 American National Standard (ANS). Further details of the development process are avai
6、lable in the TIA Engineering Manual, located at http:/www.tiaonline.org/standards/sfg/engineering_manual.cfm TIA Documents shall be reviewed on a five year cycle by the formulating Committee and a decision made on whether to reaffirm, revise, withdraw, or proceed to develop an American National Stan
7、dard on this subject. Suggestions for revision should be directed to: Standards that a certain course of action is preferred but not necessarily required; or that (in the negative 23 form) a certain possibility or course of action is discouraged but not prohibited. “May” and “need not” indicate 24 a
8、 course of action permissible within the limits of the standard. “Can” and “cannot” are used for statements of 25 possibility and capability, whether material, physical, or causal. 26 2 THE BAUDOT CODE 27 TTY devices transmit characters using a 45.45 bps or 50 bps Baudot code. Each bit has a nominal
9、 duration of 28 22 ms for 45.45 bps and 20 ms for 50 bps. A character consists of 1 start bit, 5 data bits, and 1-2 stop bits. The 29 code uses a carrierless, binary FSK signaling scheme. A mark, or “1”, is transmitted with a 1400 Hz tone. The 30 space, or “0”, is transmitted with an 1800 Hz tone. 3
10、1 In TTY devices, the first tone of a character may be the space tone of the start bit, or a mark tone preceding the 32 start bit. Furthermore, there is a mark hold tone, which extends the length of time the stop bit is transmitted 33 following the last character from 150 ms to 300 ms. The mark hold
11、 tone is not transmitted if the character is 34 immediately followed by another character. The mark hold tone prevents the transmitting TTY device from 35 receiving its echo and mistaking it for an incoming character. This mechanism is effective for wireline calls but 36 may not be sufficient to mit
12、igate echo in wireless scenarios because of the longer delays. Refer to the TTY 37 specification for a solution to mitigating echo in wireless networks. 38 With 5 data bits, the Baudot code is only capable of 32 different characters, which is not enough to represent 39 the alphabet, numbers, and pun
13、ctuation. This problem is remedied by introducing “shift” characters that 40 Deleted: and noise suppression, all Copyright Electronic Industries Alliance Provided by IHS under license with EIANot for ResaleNo reproduction or networking permitted without license from IHS-,-,-TIA-889-A change the way
14、the receiving TTY device interprets the 5 data bits. There exists the letters shift character (31) 1 and the figures shift character (27). Following the letters shift character, the receiving TTY device interprets 2 the 5-bit code using the letters library, which consists of letters of the alphabet.
15、 The figures library contains 3 numbers, punctuation, and special characters. The TTY device automatically sends the shift character when the 4 user switches from characters in one library to the other. The consequence of this scheme is that if the shift 5 character is received incorrectly, every ch
16、aracter thereafter will be misinterpreted with the wrong library until 6 another shift character is received correctly, causing a string of character errors. 7 3 CHARACTER ERROR RATE 8 Because of the shift character, it is possible to receive the 5-bit Baudot code correctly, but print the wrong 9 ch
17、aracter if the shift key was received in error. Therefore, there exist two different measures of character error 10 rates; Printable Character Error Rate (PCER) and Total Character Error Rate (TCER) 5. PCER compares the 11 actual text sent and received without any consideration to the underlying met
18、hod of transfer, which involved 12 conversion to and from Baudot with the insertion of shift state characters. TCER recognizes the Baudot 13 character set and the insertion of shift characters by comparing the 5-bit Baudot code directly. 14 For both PCER and TCER, the character error rate is compute
19、d by counting the number of missed and changed 15 characters. The number of characters in the reference file is considered the total number of characters. PCER 16 = (missed + changed)/total for the printable characters and TCER = (missed + changed)/total for the 5-bit 17 Baudot code character set. T
20、he number of added characters is not considered in the character error rate. 18 The TTY Forum has set a goal for the character error rate to be less than 1%, however, there was no agreement 19 as to whether that would be TCER or PCER. The TTY solution adopted for CDMA is capable of far exceeding 20
21、this goal. For the purposes of this minimum performance requirement, only the TCER shall be used because it 21 does not depend on the content of the TTY sequences. Furthermore, the TCER shall be calculated using the 22 tools provided with this specification in conjunction with the UNIX diff command.
22、 See Section 5 for a 23 description of the tools. 24 4 TTY/TDD TEST CASES 25 This section describes the test cases that are exercised in this minimum performance specification. Compliance 26 with these requirements shall be verified by processing the accompanying test vectors and using the 27 accomp
23、anying tools, as described in Section 6. 28 4.1 Interoperability with Reference TTY Vocoder 29 The test vocoder shall interoperate with the simulation of its corresponding reference vocoder. Packets from 30 the test encoder shall interoperate with the reference decoder, and vice versa. 31 4.2 Intero
24、perability with Reference Non-TTY Capable Vocoder 32 The test vocoder shall interoperate with legacy, non-TTY capable vocoders. Packets from the test encoder shall 33 interoperate with the reference decoder, and vice versa, for TTY and non-TTY inputs. Performance with the 34 test vocoder and non-TTY
25、 vocoders shall be comparable to the performance between two non-TTY capable 35 phones. 36 4.3 Input Level 37 The test vocoder shall meet the performance requirements for detection and regeneration when the input signal 38 level ranges from 45 dBm to 5 dBm. 39 Copyright Electronic Industries Allianc
26、e Provided by IHS under license with EIANot for ResaleNo reproduction or networking permitted without license from IHS-,-,-TIA-889-A 4-3 4.4 Bit Duration 1 The test vocoder shall meet the performance requirements for detection and regeneration when the input bit 2 durations are in the range 22 ms 0.
27、4 ms for 45.45 bps and 20 ms 0.4 ms for 50 bps. 3 4.5 Stop Bit 4 The test vocoder shall meet the performance requirements for detection and regeneration with stop bits ranging 5 in duration from 1 to 2 bits, where the bits may be in the range 22 ms 0.4 ms for 45.45 bps and 20 ms 0.4 6 ms for 50 bps.
28、 7 4.6 Phase Changes 8 The performance of the test vocoder shall be independent of the inputs phase changes between TTY bits. 9 These phase changes include: 10 Zero phase 11 Random phase 12 4.7 Frequency Deviations 13 The test vocoder shall meet the performance requirements for detection and regener
29、ation when the mark tone 14 (1400 Hz) and the space tone (1800 Hz) deviate from their nominal values by 4.0%. 15 4.8 Typing Mode 16 For most applications, text telephone conversations are conducted live, with manual typists. The TTY call is 17 characterized by bursts of typing followed by pauses. Th
30、is is referred to as manual mode. Some TTY devices 18 are capable of transmitting stored conversations, or files. In this case, a long, uninterrupted stream of TTY 19 characters characterizes the TTY call. This is referred to as streaming mode. The test vocoder shall be capable 20 of supporting both
31、 typing modes. 21 4.9 Preamble 22 In TTY devices, the first tone of a character may be the space tone of the start bit, or a mark tone preceding the 23 start bit, which is referred to as the preamble. The test vocoder shall meet the performance requirements for 24 detection and regeneration whether
32、or not the preamble is present. 25 4.10 Mark Hold Tone 26 The mark hold time defines an additional period of time during which the TTY transmits a mark hold tone 27 (1400 Hz) following the last character transmitted. The mark hold tone is not transmitted between each 28 character if the character is
33、 followed immediately by another character. The test vocoder shall regenerate a 29 mark hold tone for 300 ms or until the next character is regenerated, whichever is less. The test vocoder shall 30 meet the performance requirements for detection and regeneration: 31 When characters are followed by t
34、he mark hold tone. 32 When characters are followed by silence. 33 Copyright Electronic Industries Alliance Provided by IHS under license with EIANot for ResaleNo reproduction or networking permitted without license from IHS-,-,-TIA-889-A 4.11 Voice Carryover/ Hearing Carryover (VCO/HCO) 1 When the T
35、TY algorithm is enabled, the test vocoder shall be capable of supporting VCO/HCO, without any 2 further interaction from the network or the end-users. That is to say that the TTY algorithm shall be capable of 3 switching between TTY and non-TTY inputs freely within a call without degrading its TTY p
36、erformance. 4 4.12 Tandem 5 The test vocoder shall be capable of tandeming with itself and with the reference vocoder for TTY calls. 6 4.13 Variable Rates 7 The test vocoder shall be capable of decoding TTY information from the compressed speech packets for all 8 rates that support TTY. That is to s
37、ay, that IS-127 EVRC and IS-733 13K vocoders shall decode TTY 9 information from both full rate and half rate frames. IS-893 SMV shall decode TTY information from half rate 10 data packets only. The rates used in future vocoders are to be determined. 11 4.14 Reduced Rate Modes 12 There are certain s
38、cenarios where the vocoder is requested to reduce its rate artificially in order to satisfy 13 network requirements. An example of this is dim-and-burst signaling or blank-and-burst signaling. During a 14 TTY call, the networks request for reduced-rate modes shall take precedence and the TTY informa
39、tion shall be 15 provided where possible. For example, if the test vocoder is requested to supply half rate packets during a TTY 16 transmission, the test vocoder shall produce half rate packets as requested with the TTY information embedded. 17 TTY transmission is not expected in cases where the sp
40、eech path is muted, as is the case in blank-and-burst 18 signaling. In these cases, the TTY information shall be recovered as best as possible, with the understanding 19 that character errors may occur. As a direct consequence, to achieve the required performance goals of this 20 specification, all
41、other tests in this document do not employ or require blank-and-burst signaling as part of a test 21 vector set. 22 4.15 False Alarms 23 The test vocoder shall be robust to false alarms. False alarms can occur in two places, the TTY encoder and the 24 TTY decoder. The TTY encoder may falsely label s
42、peech or tones as TTY and send TTY_SILENCE messages 25 or TTY characters to the decoder. Likewise, the TTY decoder may misinterpret non-TTY packets as TTY 26 information and either mute its output or replace its output with regenerated tones. For the purposes of this 27 document, a TTY false alarm i
43、s defined as an action by the TTY algorithm, either the TTY encoder or the TTY 28 decoder, that causes muting or tones being mistakenly regenerated at the decoders output. 29 In order to make the reference vocoder robust to false alarms, more stringent rules are applied to the first 30 character of
44、a call. The rules are relaxed for all subsequent characters within the same call. As a result of the 31 more stringent rules, there may be a character error at the beginning of the call as a result. The test vocoder 32 may also exhibit similar performance. 33 4.16 Slew 34 Certain TTY devices exhibit
45、 a decaying DC bias, or slew, when generating a character after a period of silence. 35 The test vocoder shall meet or exceed the performance requirements for detection and regeneration in the 36 presence of slew. 37 Deleted: ,Deleted: , and IS-893 SMV Copyright Electronic Industries Alliance Provid
46、ed by IHS under license with EIANot for ResaleNo reproduction or networking permitted without license from IHS-,-,-TIA-889-A 5-5 4.17 Impaired Channels 1 The test vocoder shall not degrade its TTY performance when operating in simulated impaired channel 2 environments with frame error rates from 0%
47、to 6%. 3 4.18 Worst Case Deviations 4 The test vocoder shall not degrade its TTY performance when operating in a combination of extreme deviations 5 for bit duration, stop bit length, frequency, slew, and phase changes and regardless of the typing mode, 6 preamble, or mark hold tone. 7 5 TEST TOOLS
48、8 5.1 ttygen 9 The tool ttygen generates random characters according to 5. The parameters for generating the Baudot tones 10 can be varied from the command line so that the range of parameters specified in this document can be tested. 11 Calling ttygen without parameters prints its help message. The
49、 parameters are specified as follows: 12 13 Usage: ttygen options 14 -a 15 -b 16 -c 17 -l 18 -s 19 -f 20 (or “+-“ for random within range) 21 -t 22 (or “+-“ for random within range) 23 -p 24 0: continuous (default), 25 1: zero phase, 26 2: random phase, 27 3: max. discontinuous 28 -g 29 -h 30 -N 31 -r 32 -x 33 5.2 ttyrcv 34 The tool ttyrcv receives a PCM stream, and detects TTY/TDD Baudot characters. The 5-bit decimal value 35 corresponding to the detected character is p
