Vertical Handoff In Heterogeneous Wireless Networks Based On Network Utility

With the rapid development of wireless communication technology and the spread ofthe all-IP technology, the next-generation wireless networks will be a heterogeneous wirelessnetworks fusion system based on the all-IP technology. How to achieve a smooth handoff ofmobile terminals as well as an effective sharing of resources between the networks under theenvironment of multiple wireless access networks co-exist, has been a problem for the issueof radio resource management in heterogeneous wireless networks. Therefore, as the keytechnologies for the radio resource management for heterogeneous wireless networks,vertical handoff algorithm and load balancing strategy, have become research topics ofcurrent concern.In this paper, aiming at solving the problem that the vertical handoff algorithms inexistence haven’t taken full account of the different requirements of network bandwidth andtime delay between the different users in heterogeneous wireless networks, we put forward amulti-weighting of the vertical handoff algorithm (MWVH) based on the user consumptionhabits, which regarded the expectations, variances and frequencies of the.real-time and thenon real-time services as the choice basis of the network bandwidth and connection delayweight factor in the algorithm. When a user requests the handoff, the mobile terminal willobtain the necessary information to calculate the profit value of each target network, andchoose the best one to put handoff into effect. Compared with the utility-based verticalhandover mechanism (UBVH), MWVH algorithm could reduce the blocking rate of newuser、call drop rate of the handoff user and the handoff times, as well as improving the usersatisfaction.To enhance the dynamic adaptability of the of MWVH algorithm, the weight factorselection method was improved by proposing a dynamic-weight vertical handoff algorithm(DWVH) based on user consumption habits in this paper. In the DWVH, a consumptionscale composed of the expectations and frequencies of the user’s consumptions was definedto compute the weight factor in the utility function in real time. In order to reduce theinfluences that the consumption volatility has on the weight factor, the algorithm correctedthe variances of the weight factor which had been calculated by comparing the variance ofreal-time service non-real-time service. The algorithm had a second correction on the resultsderived from the above calculation to meet user’s requirements of the network bandwidthand time delay, DWVH algorithm solved the problem of the limited numbers of weightfactor in MWVH algorithm, and the simulation shows that the DWVH algorithm can makegreat contributions in reducing the blocking rate of new user、call drop rate of the handoffuser and the handoff times, in the meanwhile, the user satisfaction has been improvedaccordingly.To enhance the performance of network load balancing and avoid network loadimbalances caused by vertical handoff, we converted the network load into the form of loadcost function, and proposed a price-load DWVH algorithm (LB-DWVH) in which the load cost function was considered as a loss included in the utility function of the DWVHalgorithm. The simulation shows that compared with the UBVH algorithm, the LB-DWVHalgorithm has improved the system utilization by considering the network load when a userselects vertical handoff target network.This work has a certain value on the improvement of the vertical handoff algorithm andthe design of the load balancing strategy in heterogeneous wireless networks.