A simple near-orthogonal code is used as frequency-hopping patterns for the frequency-hopped multiple access systems. Extended RS code is used as channel coding to deplete the effects of hits from simultaneous users. Packet error probability and channel throughput for the system utilizing the near-orthogonal code are evaluated and compared to the corresponding values obtained from the system utilizing random patterns. Results show that the former can provide substantial improvement over the latter. In our illustrated examples, we also show that under the constraint of packet error probability PE 10-2, the maximum achievable number of users with most (n,k) RS codes of interest is less than the number of distinct codewords in the near-orthogonal code. Thus, the number of codewords of the near-orthogonal code is large enough to support the practical application.

This paper presents the performance of FH/MFSK systems, which exploit silent gaps in speech to accommodate more users, over Rayleigh fading channels. Two kinds of receivers are considered: one uses a threshold on the received signal strength to declare whether the signals were present or not, and the other is assumed to have perfect transmitter-state information obtained from using additional bandwidth. Results show that, if the codeword dropping and codeword error are assumed to be equally costly, the former can achieve slightly better performance than the latter in the decoding error probability. This finding suggests that, for the system to exploit silent gaps in speech, it is advantageous for the receiver to use a threshold to declare whether signals were present or not instead of relying on the transmitter-state information.