Fuzzy/neural congestion control for DS-CDMA/FRMA cellular systems

We propose a fuzzy/neural congestion controller for DS-CDMA/FRMA cellular systems. The congestion controller adopts fuzzy logic and neural network techniques to regulate the transmission permission probability of contention users, according to the predicted interference and indicated performance. It contains a pipeline recurrent neural network (PRNN) interference predictor, a fuzzy performance indicator, and a fuzzy permission probability controller. Simulation results show that the fuzzy/neural congestion control for DS-CDMA/FRMA is more adaptive and has lower corruption ratio by 36.67% and smaller voice packet dropping ratio by 15.15%, compared to the channel access function for DS-CDMA/PRMA proposed by Brand and Aghvami (see IEEE J. Select. Areas Commun., vol.14, no.9, p.1698-1707, 1996)