TIA ANSI/TIA/EIA-136-410 PDF

Preface

This document gives a description of the ACELP speech and channel codec for a TIA Enhanced Full Rate Codec. The codec consists of a 7.4 kbitls ACELP speech codec and two channel codecs (Forward Error Correction): a 5.6 kbitls channel codec and a 6.5 kbitls channel codec.

Section 1 gives a general description of the speech coder and sections 2 and 3 give more detailed descriptions of the speech encoder and the speech decoder, respectively. Section 4 gives a detailed description of the two channel encoders. The channel decoder in section 5 is not considered as part of this standard. The bad frame masking described in section 5.1 is provided as an example solution; other solutions are allowed. Section 6 gives a detailed description of the different aspects of Discontinuous Transmission (DTX). The bad frame masking described in section 6.4.4 is provided as an example solution; other solutions are allowed.

This standard consists of this document and a soft copy distribution comprising on a CD ROM. These contain a bit-exact description in the form of a fixed point ANSI C code and a number of test sequences. The C code defines the speech codec and the channel encoder with a set of fixed point basic operators. The test vectors are described in Annex A. A detailed description of the contents and formats of the soft copy distribution is given in Annex B. The details of the instability protection for the speech codec are given in Annex C. Annex D describes a reference design of a voice activity detector (VAD) which may be used to implement the DTX mode of the speech encoder in the mobile station (MS).

MS implementations shall be bit-exact. Bit-exact implementations are subject to the test vectors specified in this standard. Base station implementations are not required to be bitexact. However, non bit-exact validation is subject to the minimum performance requirements for the Enhanced Full Rate codec contained in TIAEIA IS-686.

In the case of discrepancy between the high level and the bit-exact descriptions, the description in the bit-exact C program will prevail.