A multiple user variable rate speech coding for packet transmission systems

In an effort to reduce queuing delay and increase the number of channels in a given bandwidth, a multiple-user variable-rate hybrid constant factor incremental data modulator is presented. The proposed system has a statistical multiplexing scheme incorporated with a packet transmission concept. Only the speech is encoded, while the silence portions are transmitted in header bits of each packet generated by a packet assembler. Computer simulations have been performed for real speech. The results indicate that the new system yields a performance improvement of about 8 dB over the conventional (single-user) systems.<
>