1、CCITT RECMN*G*726 90 W 4862593 0561826 8 W INTERNATIONAL TELECOMMUNICATION UNION CCITT THE INTERNATIONAL TELEGRAPH AND TELEPHONE CONSULTATIVE COMMITTEE GENERAL ASPECTS OF DIGITAL TRANSMISSION SYSTEMS; TERMINAL EQUIPMENTS G.726 40, 32, 24, 16 kbitls ADAPTIVE DIFFERENTIAL PULSE CODE MODULATION (ADPCM)
2、 Recommendation G.726 T Geneva, 1990 CCITT RECMN*G*72b 90 4862591 0561827 T INTERNATIONAL TELECOMMUNICATION UNION CCITT THE INTERNATIONAL TELEGRAPH AND TELEPHONE CONSULTATIVE COMMITTEE GENERAL ASPECTS OF DIGITAL TRANSMISSION SYSTEMS; TERMINAL EQUIPMENTS G.726 40, 32, 24, 16 kbit/s ADAPTIVE DIFFERENT
3、IAL PULSE CODE MODULATION (ADPCM) Recommendation G.726 Geneva, 1990 CCITT RECMN*Ge72b 90 E 48b259L 05bL828 L E FOREWORD The CCITT (the International Telegraph and Telephone Consultative Committee) is a permanent organ of the International Telecommunication Union (ITU). CCITT is responsible for study
4、ing technical, operating and tariff questions and issuing Recommendations on them with a view to standardizing telecommunications on a worldwide basis. The Plenary Assembly of CCITT which meets every four years, establishes the topics for study and approves Recommendations prepared by its Study Grou
5、ps. The approval of Recommendations by the members of CCITT between Plenary Assemblies is covered by the procedure laid down in CCITT Resolution No. 2 (Melbourne, 1988). Recommendation G.726 was prepared by Study Group XV and was approved under the Resolution No. 2 procedure on the 14 of December 19
6、90. CCIIT NOTE In this Recommendation, the expression “Administration“ is used for conciseness to indicate both a telecommunication Administration and a recognized private operating agency. O ITU 1990 All rights reserved. No part of this publication may be reproduced or utilized in any form or by an
7、y means, electronic or mechanical, including photocopying and microfilm, without permission in writing from the ITU. CCITT RECMN*G*726 90 4862.591 0561827 3 ,R Recommendation G.726 40, 32, W, 16 kbitls ADAPTIVE DIFFERENTIAL PULSE CODE MODULATION (ADPCM) 1 General The characteristics below are recomm
8、ended for the conversion of a 64 kbit/s A-law or p-law pulse code modulation (PCM) channel to and from a 40, 32, 24 or 16 kbit/s channel. The conversion is applied to the PCM bit stream using an ADPCM transcoding technique. The relationship between the voice frequency signals and the PCM encoding/de
9、coding laws is fully specified in Recommendation G.711. The principal application of 24 and 16 kbit/s channels is for overload channels carrying voice in Digital Circuit Multiplication Equipment (DCME). The principal application of 40 kbit/s channels is to carry data modem signals in DCME, especiall
10、y for modems operating at greater than 4800 kbit/s. Sections 1.1 and 1.2 of this Recommendation provide an outline description of the ADPCM transcoding algorithm, $9 2 and 3 provide the principles and functional descriptions of the ADPCM encoding and decoding algorithms respectively, whilst 9 4 is t
11、he precise specification for the algorithm computations. Networking aspects and digital test sequences are addressed in Appendices I and II, respectively, to this Recommendation. Simplified block diagrams of both the ADPCM encoder and decoder are shown in Figure 1/G.726. In 9 4, each sub-block in th
12、e encoder and decoder is precisely defined using one particular logical sequence. If other methods of computation are used, extreme care should be taken to ensure that they yield exactly the same value for the output processing variables. Any further departures from the processes detailed in Q 4 wil
13、l incur performance penalties which may be severe. Note 1 - Prior to the definition of this Recommendation, other ADPCM algorithms of performance similar to the 40 kbit/s algorithm specified here have been incorporated in DCME designs and used in telecommunications networks. These algorithms may be
14、considered by bilateral agreement for limited DCME applications under certain circumstances. Technical descriptions providing information on two such algorithm approaches can be found in COM XVIII No. 101 and COM XVIII No. 102 of the 1984-1988 Study Period. Note 2 - The assignment of 16,24, 32 and 4
15、0 kbit/s DCME channels and the associated selection of coding rates are beyond the scope of this Recommendation; see, for example, Recommendation G.763 (revised, 1990). Note 3 - Signalling and multiplexing considerations are beyond the scope of this Recommendation; see, for example, Recommendations
16、G.761 and G.763 (revised, 1990). 1 This Recommendation completely replaces the text of Recommendations G.721 and G.723 published in Volume m.4 of the Blue Book. It should be noted that systems designed in accordance with the present Recommendation will be compatible with systems designed in accordan
17、ce with the Blue Book version. Recommendation G.726 1 Difference signal 64 kbits , Convert to PCM input output quantizer uniform PCM b Adaptative ADPCM b 4 I 4- -: t Recon- Adaptive predictor structed signa Inverse adaptive quantizer U Quantized U difference signal a) Encoder ADPCM input Quantized I
18、 l Inverse quantizer output adjustment + b coding -+ Convert to PCM adaptive c Synchronous 64 kbit/s Signal estimate I t 4 Adaptive predictor e T1502500-90 61 Decoder FIGURE 1/G.726 Simplified block diagrams 1.1 ADPCM encoder Subsequent to the conversion of the A-law or y-law PCM input signal to uni
19、form PCM, a difference signal is obtained, by subtracting an estimate of the input signal from the input signal itself. An adaptive 31-, 15-, 7-, or 4-level quantizer is used to assign five, four, three or two binary digits, respectively, to the value of the difference signal for transmission to the
20、 decoder. An inverse quantizer produces a quantized difference signal from these same five, four, three or two binary digits, respectively. The signal estimate is added to this quantized difference signal to produce the reconstructed version of the input signal. Both the reconstructed signal and the
21、 quantized difference signal are operated upon by an adaptive predictor which produces the estimate of the input signal, thereby completing the feedback loop. 1.2 ADPCM decoder The decoder includes a structure identical to the feedback portion of the encoder, together with a uniform PCM to A-law or
22、p-law conversion and a synchronous coding adjustment. The synchronous coding adjustment prevents cumulative distortion occurring on synchronous tandem codings (ADPCM-PCM-ADPCM, etc., digital connections) under certain conditions (see 5 3.7). The synchronous coding adjustment is achieved by adjusting
23、 the PCM output codes in a manner which attempts to eliminate quantizing distortion in the next ADPCM encoding stage. 2 Recommendation G.726 CCITT RECMN*G*72b 90 LI862593 0563833 I 2 ADPCM encoder principles Figure 2/G.726 is a block schematic of the encoder. For each variable to be described, k is
24、the sampling index and samples are taken at 125 p s intervals. A fundamental description of each block is given below in $5 2.1 to 2.8. - ADPCM Reconstructed calculator - output 3 signal Sr(k) Input PCM adaptive Adaptive signal format Difference Inverse Adaptive s(k)- a) predictor quantizer dqk) . I
25、(k) quantizer do- computation sl(kT- conversion A ta t t I I a, (k) II I - c v Quantizer tdk) detector control transition tr(k) 4 adaptation speed Tone and yk) scale factor 4 = Adaptation C 4 a, (k) Yl(k) I t Tl502SIO-90 FIGURE 2/G.726 Encoder block schematic 2.1 Input PCM format conversion This blo
26、ck converts the input signal s (k) from A-law or p-law PCM to a uniform PCM signal sl(k). 2.2 Difference signal computation 2.3 Adaptive quantizer A 31-, 15, 7- or 4-level non-uniform adaptive quantizer is used to quantize the difference signal d (k) for operating at 40, 32, 24 or 16 kbit/s, respect
27、ively. Prior to quantization, d (k) is converted to a base 2 logarithmic representation and scaled by y (k) which is computed by the scale factor adaptation block, The normalized input/output characteristic (infinite precision values) of the quantizer is given in Tables 1/G.726 through 4/G.726. Reco
28、mmendation G.726 3 CCITT RECMN*G*72b 90 4862593 056LB32 3 2.3.1 Operation at 40 kbitls Five binary digits are used to specify the quantized level representing .d (k) (four for the magnitude and one for the sign). The 5-bit quantizer output I(k) forms the 40 kbit/s output signal; I (k) takes on one o
29、f 3 1 non-zero values, I(k) is also fed to the inverse adaptive quantizer, the adaptation speed control and the quantizer scale factor adaptation blocks that operate on a 5-bit I (k) having one of 32 possible values. I (k) = O0000 is a legitimate input to these blocks when used in the decoder, due t
30、o transmission errors. TABLE 1/G.726 Quantizer normalized input/output characteristic for 40 kbit/s operation Normalized quantize input range log2 I (4 I - Y (4 4.31, + - ) 4.12,4.31) 3.91,4.12) 3.70, 3.91) 3.47, 3.70) 3.22, 3.47) 2.95, 3.22) 2.64,2.95) 2.32,2.64) 1.95,2.32) 1.54, 1.95) 1.08, 1.54)
31、0.52, 1.08) -O. 13, 0.52) -0.96, -0.13) (- 00, -0.96) 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 O Normalized quantizer output log2 I dq (4 I - Y (4 4.42 4.21 4.02 3.81 3.59 3.35 3.09 2.80 2.48 2.14 1.75 1.32 0.81 0.22 -0.52 -00 Note - In Tables 1/G.726 through 4/G.726, “I indicates that the endpoint value
32、 is included in the range, and “(” or “)” indicates that the endpoint value is excluded from the range. 2.3.2 Operation at 32 kbitls Four binary digits are used to specify the quantized level representing d (L) (three for the magnitude and one for the sign). The 4-bit quantizer output I (k) forms th
33、e 32 kbit/s output signal; it is also fed to the inverse adaptive quantizer, the adaptation speed control and the quantizer scale factor adaptation blocks. I (k) = O000 is a legitimate input to these blocks when used in the decoder, due to transmission errors. 4 Recommendation G.726 CCITT RECMN*G*72
34、b 90 4862591 0561833 5 = TABLE 2lG.726 Quantizer normalized input/output characteristic for 32 kbitls operation Normalized quantizer Normalized quantizer input range output I I (M I log:! I d (k) I - Y (4 log2 I dq (4 I - Y (M 3.12, + - ) 1 .O5 2 0.62, 1.38) 1.66 3 1.38, 1.91) 2.13 4 1.91,2.34) 2.52
35、 5 2.34, 2.72) 2.91 6 2.72, 3.12) 3.32 7 -0.98, 0.62) 0.031 1 (- -,-0.98) -00 O 2.3.3 Operatio n at 24 kbitls Three binary digits are used to specify the quantized level representing d (k) (two for the magnitude and one for the sign). The 3-bit quantizer output I (k) forms the 24 kbids output signal
36、, where I (k) takes on one of sevel non- zero values. I (k) is also fed to the inverse adaptive quantizer, the adaptation speed control and the quantizer scale factor adaptation blocks, each of which is modified to operate on a 3-bit I(k) having any of the eight possible values. I (k) = O00 is a leg
37、itimate input to these blocks when used in the decoder, due to transmission errors. TABLE 3lG.726 Quantizer normalized input/output characteristic for 24 kbit/s operation Normalized quantizer input range log2 I (4 I - Y (4 2.58, + - ) 1.70, 2.58) 0.06, 1.70) (- W,-0.06) I 2.3.4 Operation at 16 kbitl
38、s Normalized quantizer I I (4 I output log2 I (4 I - Y (k) 3 2 2.91 -00 O 1 .O5 1 2.13 Two binary digits are used to specify the quantized level representing d (k) (one for the magnitude and one for the sign). The 2-bit quantizer output I (k) forms the 16 kbit/s output signal; it is also fed to the
39、inverse adaptive quantizer, the adaptation speed control and the quantizer scale factor adaptation blocks. Recommendation G.726 5 CCITT RECMN*G-726 90 m 48b2573 0563834 7 m TABLE 4jG.726 Quantizer normalized input/output characteristic for 16 kbit/s operation Normalized quantizer Normdized quantizer
40、 input range output 1 I (kj I log2 I d (4 I - Y (4 log2 I dq (k) I - Y (k) 2.04, + - ) 2.85 1 (-W, -2.04) 0.91 O Unlike the quantizers described in Q 2.3.1 for operation at 40 kbit/s, in Q 2.3.2 for operation at 32 kbit/s and in Q 2.3.3 for operation at 24 kbit/s, the quantizer for operation at 16 k
41、bit/s is an even-level (4-level) quantizer. The even-level quantizer for the 16 kbit/s ADPCM has been selected because of its superior performance over a corresponding odd-level (3-level) quantizer. 2.4 Inverse adaptive quantizer A quantized version dq (k) of the difference signal is produced by sca
42、ling, using y (k), specific values selected from the normalized quantizing characteristic given in Tables UG.726 through 4/G.726 and then transforming the result from the logarithmic domain. 2.5 Quantizer scale factor adaptation This block computes y (k), the scaling factor for the quantizer and the
43、 inverse quantizer. The inputs are the 5-bit, 4-bit, 3-bit, 2-bit quantizer output I (k) and the adaptation speed control parameter al (k). The basic principle used in scaling the quantizer is bimodal adaptation: - fast for signals (e.g. speech) that produce difference signals with large fluctuation
44、s; - slow for signals (e.g. voiceband data, tones) that produce difference signals with small fluctuations. The speed of adaptation is controlled by a combination of fast and slow scale factors. The fast (unlocked) scale factor yu (k) is recursively computed in the base 2 logarithmic domain from the
45、 resultant logarithmic scale factor y (k): yu(k) = (1 - 2-5) y(k) +2-5 W I), (2-2) where yu (k) is limited by 1.06 5 yu (k) I 10.00. For 40 kbit/s ADPCM, the discrete function W(r is defined as follows (infinite precision values): II(k)l I 15 I 14 I 13 I 12 I 11 I 10 I 9 I 8 WI(k) I 43.50 I 33.06 I
46、27.50 I 22.38 I 17.50 I 13.69 I 11.19 I 8.81 II(k)l1716l514131211IO WI(k) I 6.25 I 3.63 I 2.56 I 2.50 I 2.44 I 1.50 I 0.88 I 0.88 6 Recommendation G.726 CCITT RECMN*G.72b 90 4862591 05b1835 9 = For 32 kbit/s ADPCM, the discrete function W(1) is defined as follows (infinite precision values): I I(k)
47、I O 1 2 3 4 5 6 7 WI(k) -0.75 1.13 2.56 4.00 7.00 12.38 22.19 70.13 For 24 kbit/s ADPCM, the discrete function W(1) is defined as follows (infinite precision values): I I(k) I O 1 2 3 WI(k) -0.25 1.88 8.56 36.38 For16 kbit/s APDCM, the discrete function W(1) is defined as follows (infinite precision
48、 values): The factor (1 - introduces finite memory into the adaptive process so that the states of the encoder and decoder converge following transmission errors. The slow (locked) scale factor y1 (k) is derived from yu (k) with a low pass-filter operation: The fast and slow scale factors are then c
49、ombined to form the resultant scale factor: r(k)=ar(k)yu(k-1)+1-ar()lyr(k-1) where O I al(/) I 1 (see Q 2.6). Recommendation G.726 7 2.6 Adaptation speed control The controlling parameter ar (k) can assume values in the range O, 11. It tends towards unity for speech signals and towards zero for voiceband data signals. It is derived from a measure of the rate-of-change of the difference signal values. and Two measures of the average magnitude of I (k) are computed: dd(k) = (1 - 2-7) dd (k - 1) + 2-7 FJ(k) For 40 kbit/s ADPCM, Fl
