Research On Key Technology Of Vertical Handoff In Heterogeneous Wireless Networks

Heterogeneous wireless networks is expected to be the trend of next generation mobile communicate network, because the needs of the user can be satisfied. The representative techology is the System Architecture Evolution (SAE) introduced by 3GPP Release8. The SAE-based heterogeneous wireless network can provide users with rich and high quality services, and realize a full IP-based network implementation plan which can integrate other non-3GPP access technologys. But in the open heterogeneous multi-layer networks, how to keep the user in state of Always Best Connected (ABC) when roaming is becoming difficult. In this context, this dissertation is devoted to researching on the handoff issues in heterogeneous wireless networks.Relying on in-depth study of the SAE network architecture and the classification of vertical handoff, this dissertation established the joint cross-layer control model in the heterogeneous wireless network firstly. In this framework, analyzed the problem of handoff trigger and access selection. Then, proposed the improvement plan. The main achievements including :1. The combined predictive handover system in the heterogeneous networks (CPHSHN) is proposed. In Forced Vertical Handoff (FVH), set the max-speed threshold in order to avoid the ping-pong effect. Meanwhile, CPHSHN focuses on two types of Received Signal Strength prediction technology. One of the RSS prediction based on Kalman filter algorithm, the other is based on the gray model and neural network combination algorithm. Respective advantages of two methods has been used, the collaboration improves forecast accuracy. Then the interactive time between the upper layers has been prepared .Simulation results show that the method can reduce the amount of decision data and improve the accuracy of switching decision.2. Toward QoS selection algorithm based on MDP (TQSAM) for heterogeneous wireless network is proposed. In Unforced Vertical Handoff (UVH), the algorithm model selected the decision attributes. After standardized, used AHP method to determine weight of each. The selected decision is given by maximum criterion solved. Through the comparison of the simulation, we can find that TQSAM is better than existing methods in field of long-term utility. In addition, it can obtain higher throughput and lower latency. The results improve the effectiveness of the network handoff.3. A handoff strategy used network security assessment and grey relational analysis based on the fuzzy theory (GRABF) is proposed. In UVH, to satisfy the user who considered the security level of access networks as chiefly factor, this strategy utilized the real-time danamic security assessment model run in the SAE network. And proposed the GRABF algorithm for the key modle. The algorithm not only avoid the weakness brought by the fuzzy values of network security, but also meet the maximum closer to the user requirements for the network performance. Simulation results show the algorithm can get high level of network security as the result of handoff target network.