Adaptive filtering for delivering multimedia services in W-CDMA cellular networks using genetic algorithms

This paper presents an adaptive filtering algorithm for the delivery of multimedia content with different quality of service (QoS) requirements in time-varying wideband code-division multiple access (W-CDMA) cellular networks. The algorithm objective is to guarantee levels of quality for two different user defined profiles (premium, or economy), while attempting to maximize the allocation of the radio channel bandwidth and maintaining the QoS requirements of downloaded traffic flows (audio, video, and data). Our proposed solution is based upon utilizing genetic algorithms (GAs) to solve a multi-objective function that selects, during a control period, the bit rates and formats for the media codecs of downloaded traffic flows and assigns the air interface channel rate and power to each user, subject to different constraints. To show the performance of the algorithm, W-CDMA systems with and without filtering capabilities are compared.