Qos enhancement in wireless VoIP networks using interactive multiple model based Kalman filter

摘要

In this paper, a QoS enhancement scheme is proposed using an Interacting Multiple Model (IMM) for wireless VoIP (WVoIP) networks. The quality of speech varies in WVoIP due to variations in channel parameters. Kalman filter can be used to enhance speech quality by predicting the channel variations. IMM consists of two models namely static and dynamic, representing the wireless VoIP mobility states and both filters work in parallel. The IMM filter estimate is computed as a weighted sum of individual filter estimates. As the computational complexity is more in IMM, a systolic array based parallel architecture is proposed to implement IMM for real time applications.