电力无线通信异构多网共存环境中的网络选择算法

A network selection algorithm in heterogeneous multi-networks coexistence of electric power wireless communications

综合考虑电力业务需求和无线专网性能因素,针对邻域网、LTE无线专网、WiMAX无线专网和230 MHz无线专网等构成的异构多网混合共存场景,提出一种网络选择算法。该算法以AHPTOPSIS算法为基础,首先构建能表征候选网络与决策属性之间关系的决策矩阵,归一化处理获得归一化决策矩阵;利用AHP算法,依据不同业务类型下决策属性与网络性能之间的相对关系定义权重,对归一化决策矩阵加权获得加权归一化决策矩阵;然后对TOPSIS算法进行改进,依据加权归一化决策矩阵确定正负理想值,利用正理想值与候选网络之间的最小距离以及负理想值与候选网络之间的最大距离确定理想网络,计算各候选网络与理想网络之间的有效距离;据此执行候选网络排序和最佳网络选择,以避免运行TOPSIS算法可能存在的排序异常。仿真结果表明,该算法能根据电力业务需求对候选网络进行正确排序和最优选择,且能在满足业务服务质量的同时,提高资源利用率。

Combining the needs of electric power business with the performance of special wireless networks, we propose a network selection algorithm for heterogeneous multi-networks coexistence including neighbor networks, LTE special wireless networks, WiMAX special wireless networks and the special wireless networks at 230M frequency. This algorithm is based on the AHP-TOPSIS algorithm. We first construct the decision matrix that can indicate the relationship between candidate networks and decision attributes. The normalized decision matrix can be obtained through normalization process. The AHP algorithm is employed to define the individual weight according to the relationship between deci- sion attributes and the network performance under different business types, and then we add weights of normalized decision matrix to obtain the weighted normalized decision matrix. We also make an improvement for the TOPSIS algorithm, determine the plus-minus ideal value according to the weighted normalized decision matrix, and use the minimum value between plus ideal value and candidate network, and the maximum value between minus ideal value and candidate network to determine the ideal network. Finally, we calculate the effective distance between each candidate network and ideal network, rank candidate networks and select the best one, which can avoid anomaly ranking probably existed in the alternative TOPSIS algorithm. Simulation results show that this algorithm can rank candidate networks correctly and select the best network. Meanwhile this algorithm can not only meet the requirement of service QoS, but also improve resource utilization efficiency.