1、 ANSI/TIA-136-410-1999 APPROVED: NOVEMBER 29, 1999 REAFFIRMED: JUNE 19, 2003 REAFFIRMED: AUGUST 14, 2013 WITHDRAWN: JUNE 12, 2015 TIA-136-410 November 1999TDMA Cellular PCS/ Radio Interface Enhanced Full-Rate VoiceCodec NOTICE TIA Engineering Standards and Publications are designed to serve the publ
2、ic interest through eliminating misunderstandings between manufacturers 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 Standards
3、 and Publications shall not in any respect preclude any member or non-member of TIA from manufacturing or selling products not conforming to such Standards and Publications. Neither shall the existence of such Standards and Publications preclude their voluntary use by Non-TIA members, either domesti
4、cally or internationally. Standards and Publications are adopted by TIA in accordance with the American National Standards Institute (ANSI) patent policy. By such action, TIA does not assume any liability to any patent owner, nor does it assume any obligation whatever to parties adopting the Standar
5、d or Publication. This Standard does not purport to address all safety problems associated with its use or all applicable regulatory requirements. It is the responsibility of the user of this Standard to establish appropriate safety and health practices and to determine the applicability of regulato
6、ry limitations before its use. Any use of trademarks in this document are for information purposes and do not constitute an endorsement by TIA or this committee of the products or services of the company. (From Project No. 3-4027.410-RF2-WD, formulated under the cognizance of the TIA TR-45 Mobile (b
7、) there is no assurance that the Document will be approved by any Committee of TIA or any other body in its present or any other form; (c) the Document may be amended, modified or changed in the standards development or any editing process. The use or practice of contents of this Document may involv
8、e the use of intellectual property rights (“IPR”), including pending or issued patents, or copyrights, owned by one or more parties. TIA makes no search or investigation for IPR. When IPR consisting of patents and published pending patent applications are claimed and called to TIAs attention, a stat
9、ement from the holder thereof is requested, all in accordance with the Manual. TIA takes no position with reference to, and disclaims any obligation to investigate or inquire into, the scope or validity of any claims of IPR. TIA will neither be a party to discussions of any licensing terms or condit
10、ions, which are instead left to the parties involved, nor will TIA opine or judge whether proposed licensing terms or conditions are reasonable or non-discriminatory. TIA does not warrant or represent that procedures or practices suggested or provided in the Manual have been complied with as respect
11、s the Document or its contents. If the Document contains one or more Normative References to a document published by another organization (“other SSO”) engaged in the formulation, development or publication of standards (whether designated as a standard, specification, recommendation or otherwise),
12、whether such reference consists of mandatory, alternate or optional elements (as defined in the TIA Procedures for American National Standards) then (i) TIA disclaims any duty or obligation to search or investigate the records of any other SSO for IPR or letters of assurance relating to any such Nor
13、mative Reference; (ii) TIAs policy of encouragement of voluntary disclosure (see TIA Procedures for American National Standards Annex C.1.2.3) of Essential Patent(s) and published pending patent applications shall apply; and (iii) Information as to claims of IPR in the records or publications of the
14、 other SSO shall not constitute identification to TIA of a claim of Essential Patent(s) or published pending patent applications. TIA does not enforce or monitor compliance with the contents of the Document. TIA does not certify, inspect, test or otherwise investigate products, designs or services o
15、r any claims of compliance with the contents of the Document. ALL WARRANTIES, EXPRESS OR IMPLIED, ARE DISCLAIMED, INCLUDING WITHOUT LIMITATION, ANY AND ALL WARRANTIES CONCERNING THE ACCURACY OF THE CONTENTS, ITS FITNESS OR APPROPRIATENESS FOR A PARTICULAR PURPOSE OR USE, ITS MERCHANTABILITY AND ITS
16、NONINFRINGEMENT OF ANY THIRD PARTYS INTELLECTUAL PROPERTY RIGHTS. TIA EXPRESSLY DISCLAIMS ANY AND ALL RESPONSIBILITIES FOR THE ACCURACY OF THE CONTENTS AND MAKES NO REPRESENTATIONS OR WARRANTIES REGARDING THE CONTENTS COMPLIANCE WITH ANY APPLICABLE STATUTE, RULE OR REGULATION, OR THE SAFETY OR HEALT
17、H EFFECTS OF THE CONTENTS OR ANY PRODUCT OR SERVICE REFERRED TO IN THE DOCUMENT OR PRODUCED OR RENDERED TO COMPLY WITH THE CONTENTS. TIA SHALL NOT BE LIABLE FOR ANY AND ALL DAMAGES, DIRECT OR INDIRECT, ARISING FROM OR RELATING TO ANY USE OF THE CONTENTS CONTAINED HEREIN, INCLUDING WITHOUT LIMITATION
18、 ANY AND ALL INDIRECT, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES (INCLUDING DAMAGES FOR LOSS OF BUSINESS, LOSS OF PROFITS, LITIGATION, OR THE LIKE), WHETHER BASED UPON BREACH OF CONTRACT, BREACH OF WARRANTY, TORT (INCLUDING NEGLIGENCE), PRODUCT LIABILITY OR OTHERWISE, EVEN IF ADVISED OF THE POSSI
19、BILITY OF SUCH DAMAGES. THE FOREGOING NEGATION OF DAMAGES IS A FUNDAMENTAL ELEMENT OF THE USE OF THE CONTENTS HEREOF, AND THESE CONTENTS WOULD NOT BE PUBLISHED BY TIA WITHOUT SUCH LIMITATIONS. TIAEIA- 136-410 Adopted Revision History for TIA/EIA-l36-410 Section(s) Description List of Figures List of
20、 Tables Section 6,7 Section 6 Figure 5 Figure 6 TSB77, December 1996 Added new figures 5 and 6 to the list. Renamed tables 6.1, 6.2 and 6.3 as 7.1, 7.2 and 7.3, respectively. Added new tables 6.1 and 6.2 to the list. Sections 6 and 7 renamed as sections 7 and 8, respectively. New section 6 inserted
21、to give a detailed description of the different aspects of discontinuous transmission (DTX) and comfort noise generation (CNG). Added new figure “Normal hangover procedure (NELAPSED 2 24)”. Added new figure “Handling of short speech bursts (NELAPSED : 13-bit right-justified AccA = AccA : 16-bit line
22、ar sample AccA = AccA + 0x0004 : Add rounding bit AccA = AccA Tl), n=O ,., LI-1, 2z(n - L,) 4L2 -1 n=L, ., L,+L2-1 where the values L1 =200 and L2 =40 areused. 7 The autocorrelations of the windowed speech s (n), n = O,. . . ,239 , are computed by 8 23 9 r(k) = cs (n)s(n- k) , k = O ,., 10, n=k 9 an
23、d a 60 Hz bandwidth expansion is used by lag windowing the autocorrelations using the window 2 10 11 wla,(i) = exp - ; - 2TiI2 , i = 1, . 10, 12 where fo = 60 Hz is the bandwidth expansion and fs = 8000 Hz is the sampling frequency. Further, r(0) is multiplied by the white noise correction factor 1.
24、0001 which is equivalent to adding a noise floor at -40 dl3. 13 14 15 2.2.2 Levinson-Durbin algorithm 16 The modified autocorrelations r (O) = 1.0001 r(0) and r (k) = r(k)w (k), k = 1, . 10, are used to obtain the LP filter coefficients lag 17 18 ak, k = 1, ., 10, by solving the set of equations. 19
25、 10 c akr (li - kl) = -r (i) , i = 1 ,.JO. k=l 20 The set of equations in (2.5) is solved using the Levinson-Durbin algorithm 3. This algorithm uses the following recursion: 21 22 6 TA/EIA-136-410 1 2 6 10 11 12 17 18 19 20 21 E(0) = r (O) For i=l to 10 do ki = - r (i) + xi: arr (i - j) / E(i - i) j
26、=1 J 1 a/i) = ki Forj=l to i-1 do J -j E(i) = (1 - k?)E(i - i) u(i) - (i-1) + kiaiL;) The final solution is given as aj = j = 1,. . . ,IO . The LP filter coefficients are converted to the LSP representation 4 for quantization and interpolation purposes. The conversions to the LSP domain and back to
27、the LP filter domain are described in the next section. 2.2.3 LP to LSP conversion The LP filter coefficients ak, k = 1, ., 10, are converted to the LSP representation for quantization and interpolation purposes. For a 10th order LP filter, the LSPs are defined as the roots of the sum and difference
28、 polynomials fi (z) = A(z) + z-“A(z-) (2.6) and respectively. The polynomials fi (z) and fi (z) are symmetric and antisymmetric, respectively. It can be proven that all roots of these polynomials are on the unit circle and they alternate each other 5. fi (z) has a root z = -1 (u = z) and f2 (z) has
29、a root z = 1 (u = O) . To eliminate these two roots, we define the new polynomials and Each polynomial has 5 conjugate roots on the unit circle (ekiwi), therefore, the polynomials can be written as 7 TIAEIA-136-410 1 2 3 10 11 12 13 14 15 16 17 18 19 20 25 and F,(z) = n(1- 2q,z- + z-2) i=2,4, ., 10
30、(2.1 O) (2.1 1) where 4i = cos(o,) with oi being the line spectral frequencies (LSF) and they satisfy the ordering property O 0.85Ri+1. This procedure of dividing the delay range into 3 sections and favoring the lower sections is used to avoid choosing pitch multiples. 11 TIAEIA-136-410 i 2.4 Impuls
31、e response computation 2 The impulse response, h(n), of the weighted synthesis filter This impulse 3 H(z)W(z) = A(z /yl) / A(z)A(z / y, is computed each subframe. 4 5 response is needed for the search of adaptive and fixed codebooks. The impulse response h(n) is computed by filtering the vector of c
32、oefficients of the filter A(z / ri) 6 extended by zeros through the two filters 1 / (z) and 1 / A(z / y2) 7 2.5 Target signal computation 8 9 10 11 12 13 18 19 20 21 22 23 2.6 24 25 26 27 28 29 30 31 32 The target signal for adaptive codebook search is usually computed by subtracting the zero-input
33、response of the weighted synthesis filter H(z)W(z) = A(z /yl) / LA(z)A(z / y2) from the weighted speech signal s,. (n) . This is performed on a subframe basis. An equivalent procedure for computing the target signal, which is used in this codec, is the filtering of the LP residual signal r(n) throug
34、h the combination of synthesis filter 1 / (z) and the weighting filter A(z / ri) / A(z / y2) . After determining the excitation for the subframe, the initial states of these filters are updated by filtering the difference between the LP residual and excitation. The memory update of these filters is
35、explained in Section 2.9. The residual signal r(n) which is needed for finding the target vector is also used in the adaptive codebook search to extend the past excitation buffer. This simplifies the adaptive codebook search procedure for delays less than the subframe size of 40 as will be explained
36、 in the next section. The LP residual is given by 10 r(n)= s(n)+Cis(n-i), n = O, ., 39. i=l Adaptive codebook search (2.20) Adaptive codebook search is performed on a subframe basis. It consists of performing closed loop pitch search, and then computing the adaptive codevector by interpolating the p
37、ast excitation at the selected fractional pitch lag. The adaptive codebook parameters (or pitch parameters) are the delay and gain of the pitch filter. In the adaptive codebook approach for implementing the pitch filter, the excitation is repeated for delays less than the subframe length. In the sea
38、rch stage, the excitation is extended by the LP residual to simplify the closed-loop search. In the first and third subframes, a fractional pitch delay is used with resolutions 1/3 in the range 19 84- and integers only in the range 85, 1431. For the second and fourth 3 :I 12 TA/EIA-136-410 3 d 9 10
39、11 12 13 14 15 16 17 18 23 24 25 26 27 28 29 30 31 1 2 subframes, a pitch resolution of 113 is always used in the range Ti - 5 7, Ti + 4 Ti is nearest integer to the fractional pitch lag of the previous (Ist or 3rd) subframe. , where Closed-loop pitch analysis is performed around the open-loop pitch
40、 estimates on a subframe basis. In the first (and third) subframe the range Top 13, bounded by 20 . 143, is searched. For the other subframes, closed-loop pitch analysis is performed around the integer pitch selected in the previous subframe, as described above. The pitch delay is encoded with 8 bit
41、s in the first and third subframes and the relative delay of the other subframes is encoded with 5 bits. The closed loop pitch search is performed by minimizing the mean-square weighted error between the original and synthesized speech. This is achieved by maximizing the term (2.21) where x(n) is th
42、e target signal and Yk (n) is the past filtered excitation at delay k (past excitation convolved with h(n) ). Note that the search range is limited around the open- loop pitch as explained earlier. The convolution yk(n) is computed for the first delay in the searched range, and for the other delays,
43、 it is updated using the recursive relation (2.22) where u(n), n = -( 143+11), ., 39, is the excitation buffer. Note that in search stage, the samples u(n), n = O, ., 39 , are not known, and they are needed for pitch delays less than 40. To simplify the search, the LP residual is copied to u(n) in o
44、rder to make the relation in Equation (2.22) valid for all delays. Once the optimum integer pitch delay is determined, the fractions from 2 to $ with a step of f around that integer are tested. The fractional pitch search is performed by interpolating the normalized correlation in Equation (2.21) an
45、d searching for its maximum. Once the fractional pitch lag is determined, the adaptive codebook vector v(n) is computed by interpolating the past excitation signal u(n) at the given phase (fraction). The interpolation is performed using two FIR filters (Hamming windowed sinc functions); one for inte
46、rpolating the term in Equation (2.21) with the sinc truncated at I11 and the other for interpolating the past excitation with the sinc truncated at f29. The filters have their cut-off frequency (-3 dl3) at 3600 Hz in the oversampled domain. 32 The adaptive codebook gain is then found by 13 TIAEIA-13
47、6-410 1 2 3 (2.23) where y(n) = v(n)*h(n) is the filtered adaptive codebook vector (zero-state response of H(z)W(z) to v(n). 4 2.7 Algebraic codebook structure and search 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 The codebook structure is based on interleaved single-pulse permutation (ISPP) des
48、ign. In this codebook, the innovation vector contains 4 non-zero pulses. All pulses can have the amplitudes +l or -1. The 40 positions in a subframe are divided into 4 tracks, where each track contains one pulse, as shown in Table 2.1. Table 2.1: Potential positions of individual pulses in the algeb
49、raic codebook. I Pulse I positions I I I io I O, 5, 10, 15, 20,25, 30, 35 I I il I 1, 6, 11, 16, 21, 26, 31, 36 I I i2 I 2, 7, 12, 17, 22, 27, 32, 37 I I4 3, 8, 13, 18,23,28, 33, 38 4, 9, 14, 19,24,29, 34, 39 The first three pulse positions are coded with 3 bits and the fourth pulse position with 4 bits, and the sign of the each pulse is encoded with 1 bit. This makes total of 17 bits per sub frame. The algebraic codebook is searched by minimizing the mean square error between the weighted input speech and the weighted synthesis speec
