Effects of PRMA on objective speech quality

When transmitting 32 kbit/s adaptive differential pulse (ADPCM) speech using Reed–Solomon error correction coding and 16 level quadrature amplitude modulation (16-QAM), our 20 slot packet reservation multiple access (PRMA) assisted cordless telecommunications (CT) scheme supported 36–38 speech users with negligible objective and subjective speech degradation. The average number of users per slot was nearly doubled due to deploying PRMA and toll quality speech was transmitted in a user bandwidth ∼ 11.6 kHz. For a channel signal-to-noise ratio (SNR) in excess of 25 dB, a Rayleigh fading channel and mobile speeds above 2 mph the speech segmental SNR degradation was less than 0.3 dB.