The Wireless Vertical Handoff Decision Algorithm with Parameter Prediction

The vertical handoff is highly time-sensitive in the heterogeneous wireless environment. The most available handoff decision methods are based on a few network parameters, and so the prediction of network parameters can intensely improve the efficiency of the vertical handoff. This paper focuses on improving the prediction model based on cost function. We first propose an adaptive network parameter prediction algorithm, which establishes the prediction function for each network parameter and computes the network´s QoS. Then the handoff process based on network parameter prediction is present to guide the handoff execution. We prove theoretically that the proposed prediction algorithm covers the available methods. We find the proposed algorithm can keep handoff stability. The handoff decisions based on real value and prediction value are compared in the WLAN inter-network environment. The results show that the proposed prediction model can eliminate the unnecessary handoffs which caused by the performance shake of the access network, can predict evolution trends of the network parameters in the case of performance gradual change and make the terminal connect to the QoS-ensured network as soon as possible.