Research On Improved Vertical Handover Algorithm In UMTS-WLAN Convergence Network

In order to satisfy user’s demand on Quality of Service(QoS), and taking full advantage of the complementary nature of different networks to achieve integration between networks is the future trend of network, it has also become a hot topic. And vertical handover algorithm is one of the key technologies of heterogeneous network handover. The main research work of this paper is to achieve integration between UMTS(Universal Mobile Telecommunications System) and WLAN networks, whose based on different access technologies, and design corresponding effective vertical handover algorithm to solve different problems by according to the business needs of users, and the performance of the proposed algorithm by simulation can demonstrate that the algorithm can effectively improve the performance of handover. The main research contents are as follows:(1) In this paper, On the basis of Analytic Hierachy Process(AHP) which is proposed in literature [4], When constructing judgment matrix of AHP, It contain a lot of subjectivity due to the influence of decision makes, this will affect the switching performance of objectivity.Considering and basing this question, We propose a vertical handover algorithm which combine Analytic Hierarchy Process and entropy method. The algorithm is based on a cost function, we choose network received signal strength(RSS), bandwidth, user’ preference and price as the property parameters of network, then we can get weight of those parameters by using AHP and entropy method, since entropy method can objectively measure the difference between networks,after combining with AHP it can be more objective to balance weight values of various types of attribute parameters, and then switching target network according to the value of the cost function. In the handover decision state we introduce switching delay to reduce unnecessary frequent switching. Simulation results show that, with compared to the vertical handover algorithm based AHP proposed in literature [4], The algorithm we propose in this paper, without affecting the mobile terminal business case, the switching frequency reduced by 3 times. Our algorithm can reduce unnecessary switching and reducing the ping-pong effect.(2) This paper proposes vertical handover algorithm by using pre-judgment, which is based on utility function. Taking into account that different networks can provide different maximum bandwidth, different maximum transmit power and different receiver threshold values, it will affect the validity and fairness of the switching decision. We propose normalizing signalstrengths(RSS) and bandwidth, then use utility function value to determine handover target network more effectively. Since switching delay in the sentencing stage is a mandatory constraint to handover network, it can’t be accurately determined switching. We introduce differential prediction algorithm in the determination process of vertical handover to predict the value of the next time, and constraint the prediction value of the next time to further accurately determine whether to perform handover network. If the destination network and the predicted value of the current network signal strength are met setting threshold(UMTS threshold-42 db,WLAN threshold-40db), then decide to switch. Simulation results show, with respect to the RSS algorithm and traditional utility function, the algorithm we proposed can reduce the number of switching 10 and 2 times, it proved that switching performance is improved.(3) Since the introduction of the pre-decision algorithm is based on the model of uniform motion of mobile terminal. The model cannot reflect some certain realities, and some business types(such as voice services) require high real-time requirement, and high speed of mobile devices can cause frequent switching, even cause serious drop rate. In order to solve the problems of those proposed, we propose vertical handover algorithm by using pre-judgment based on the acceleration. The algorithm is based on the speed state of the mobile device and traffic state to select the optimal network as the switching target network, we set the threshold speed 15m/s. With compared to the RSS algorithm, the traditional utility function and handover decision algorithm by using pre-judgment, the results show that, switching times were reduced by 13 times, 3 times and 1 time; handover call dropping rates are reduced by 4.2%, 2.1% and1.2%; new call blocking rate decreased by 4.2%, 2.7% and 1.8%. The three performance indicators confirm the superiority of the performance of proposed algorithm.