Mobile Wireless Communications in .NET Display Code 128 Code Set C in .NET Mobile Wireless Communications

Mobile Wireless Communications using barcode creator for .net vs 2010 control to generate, create code-128c image in .net vs 2010 applications. International Standard Serial Numbers s(n) input signal samples weighting filter W(z). e(n) e2(n) error min. L I H weighted synthesis filter H(z) ex(n): excitation long-term filter state L codebook 1 I codebook 2 H (a) coder long-term L filter state codebook 1 ex(n) A(z) output speech codebook H 2 (b) decoder Figure 8.44 IS-136 codec: VSELP system Second-generation, digital, wireless systems from Fig. 8.44 how sim ilar the VSELP codec is to the basic CELP codec of Fig.

8.43. The prime difference is in the use of two stochastic codebooks, rather than one as in CELP.

The long-term lter shown is represented by the single-pole lter of (8.10). Its pitch parameter (also called the lag parameter) is given by the parameter L indicated in Fig.

8.44, while the gain parameter in Fig. 8.

44 corresponds to the parameter b in (8.10). L ranges in value from 20 to 146, in units of the sampling interval of 125 msec, corresponding to 127 possible values.

It can thus be represented by seven bits. The weighted synthesis lter H(z) of the encoder and the lter A(z) both correspond to the all-pole transfer function of (8.9), each with m = 10 weighting parameters labeled i , i = 1 10, as indicated in Fig.

8.44. Signal samples are weighted as well by the weighting lter W(z) shown.

This lter is structured with both zeros and poles, using the same parameters i in both the numerator (zeros) and denominator (poles) of W(z). Codewords in each codebook are given by 40-element random vectors, de ned in terms of seven 40-element random basis vectors chosen for each codebook. There are thus 27 = 128 possible codewords in each codebook from which to choose an appropriate excitation vector.

Operation of the codec proceeds as follows: the error term (n) is squared and summed over 5 msec intervals, corresponding to 40 signal samples. This is then minimized over the various parameters and codewords, minimizing rst with respect to the long-term lter as the only excitation, and then moving sequentially on to the codewords, shown as parameters I and H in Fig. 8.

44. The parameters and codewords minimizing the squared error are then transmitted over the air interface between base station and mobile to the receiver decoder shown in Fig. 8.

44(b). It is again worth stressing at this point that compressed signal samples or differences in such samples are not transmitted over the air interface, as would be the case for PCM or versions of DPCM. It is the adaptively varying CELP parameters that are transmitted to the decoder.

This decoder of Fig. 8.44(b), just as the one in Fig.

8.43(b) for the CELP system, reproduces the speech model of the encoder of Fig. 8.

44(a). The parameters and codewords are actually transmitted every 20 msec or every four 5 msec intervals. Why is a 20 msec interval chosen for sending the VSELP parameters Recall from our discussion of IS-136 in Section 8.

2 that two slots per 40-msec frame, each slot carrying 260 data bits, are used to provide full-rate voice transmission (see also Fig. 8.9).

The transmission of VSELP parameters at 20 msec intervals thus provides codec adaptation twice each IS-136 frame. These parameters are sent as 159 bits every 20 msec, as noted above, corresponding to a raw voice bit rate of 7.95 kbps, the IS-136 compressed bit rate mentioned earlier.

Coding to be described below results in a coded bit rate for voice of 13 kbps, however. The 159 bits are divided among the various parameters and codewords as follows: the three gains , 1 , and 2 are allocated a total of eight bits per 5 msec sub-interval, or 32 bits for the 20 msec interval. The two codewords I and H each require seven bits per 5 msec sub-interval, as indicated above, for a total of 56 bits each 20 msec interval.

Seven bits per 5 msec sub-interval are needed to transmit the parameter L, as already noted above, for a total of 28 bits over 20 msec. Thirty-eight bits per 20 msec are used to transmit the ten lter coef cients. Finally, the energy per 20 msec frame is transmitted as well, using ve bits.

The total is thus 159 bits every 20 msec, as stated above..
Copyright © . All rights reserved.