改善用户体验的垂直切换算法

针对由超密集异构无线网络的超高动态性而引起掉话率不断增长的问题，并考虑到以往基于模糊逻辑相关垂直切换算法的较大时间开销，该文提出一种改善用户体验的垂直切换算法。首先利用5G核心访问和移动管理功能发现终端附近的所有候选网络，同时，利用自组织网络技术的环境感知能力，随时监测网络的运行状况，主动维护网络之间的邻居关系表。然后，引入动态模糊神经网络(DFNN)算法来执行切换判决，将获取到的网络参数数据作为该系统的输入，动态生成对垂直切换有效的规则库，经学习之后，计算得到输出判决值，从而为终端选择最佳接入网络。最后，仿真结果表明，该算法能够明显改善垂直切换过程中的掉话情况，降低切换失败概率。同时，与其他同类算法的时间消耗相比，该算法能够维持在一个较低的水平。

Vertical Handoff Algorithm for Improving User Experience

In order to solve the problem that the drop rate keeps increasing due to the ultra-high dynamic characteristics of ultra-dense heterogeneous wireless networks, and considering the large time cost of previous vertical handoff algorithm based on fuzzy logic correlation, a vertical handoff algorithm for improving user experience is proposed. Firstly, 5G core access and mobile management functions are used to discover all candidate networks near the terminals. At the same time, the environment awareness ability of Self-Organized Network(SON) technology is used to monitor the running status of networks at any time and maintain actively the neighbor relationship table between them. Then, the Dynamic Fuzzy Neural Network (DFNN) algorithm is introduced to execute the handover decision, and the network parameters obtained are taken as the input of the system to generate dynamically a rule base that is effective for vertical handoff. After learning, the output decision value is calculated, and the best access network for the terminal is selected. Finally, the simulation results show that the algorithm can significantly alleviate the drop of calls in the process of vertical handoff and reduce the probability of handover failure. Meanwhile, compared with other similar algorithms, it can maintain a lower time cost.