摘要

Emerging multimedia communications have introduced a scenario in which services with different quality of service (QoS) requirements and variable transmission rates coexist. It has been suggested in the literature that transmitter power control is a feasible approach to fulfilling the scenario. This paper presents a new application of an adaptive network-based fuzzy inference system (ANFIS) to power control in a multi-rate multimedia direct-sequence code-division multiple-access (DS-CDMA) system. By using fuzzy inference and taking advantage of the on-line learning ability of adaptive networks, the proposed ANFIS-assisted power control scheme precisely predicts the tendency for channel variations and thus compensates the effect of signal fading in advance. Simulation results show that the A NHS-assisted power control scheme significantly decreases the probability of unsuccessful transmission and of outage while achieving a higher average transmission rate.