In future heterogeneous wireless network environment,mobile users always want to be able to access the most appropriate network to satisfy the demand of data service.To work out the problem that exists when normalized index value serves as decision value, the utility theory is applied to multi⁃attribute algorithm,and the utility function of index under different services,terminal states and user types is given,and the normalized index value is replaced by utility value to calculate total utility of alternative network, finally the optimal network is selected by the comparison of total utility value. Simulation shows that this algorithm can better reflect the needs of users,and help choose network suitable for the user.%在未来的异构无线网络环境中，为满足数据业务的需求，移动用户总希望能够接入最适合的网络。针对将实际测量的指标值归一化后作为决策值中存在的问题，提出了将效用理论应用到多指标算法中，设计了指标在不同业务、终端状态、用户类型下的效用函数，并用指标效用替代归一化测量值作为决策值计算备选网络的总效用，最后根据总效用值的比较来选择最优网络。仿真表明，此算法能够更好地反映用户的需求，帮助用户选择适合的网络。
展开▼
