Performance prediction of GSM digital cellular speech and image transmission

The Group Special Mobile (GSM) digital cellular system provides the European countries with a standard land mobile radio (LMR) architecture. This paper illustrates how voice coded (vocoded) speech and uncompressed image transmission can be expected to degrade as a function of signal to noise ratio (SNR) and carrier to interference ratio (CIR) for a cochannel GSM interferer. Performance is predicted by Monte Carlo simulation of the system. This simulated GSM system employed (1) a Gaussian minimum shift keying (GMSK) modem with convolutional coding and Viterbi decoding, (2) a five-path urban channel impulse response with additive white Gaussian noise (AWGN), (3) a minimum mean square error (MMSE) equalizer, and (4) limiter-discriminator detection. These results demonstrate that GSM systems in an urban environment can transport useable speech and images without packet retransmission provided the SNR⩾8 dB and CIR⩾8 dB