P2P网络节点信任路径优化算法仿真

Simulation of Algorithm for P2P Network Node Trust Path Optimization

关于网络通信节点优化问题针对P2P网络系统中由于节点频繁退出易造成网络系统稳定性下降、节点信任评估较难问题,通过将差分进化算法引入P2P网络节点信任路径的粒子群进化问题求解过程,对种群的个体活性进行半协同变异,提出了半协同进化的P2P网络节点信任路径优化算法,首先利用基于粒距和权系数修正的改进粒子群算法,对P2P网络节点信任路径进行初次迭代优化求解,按照平均适应度生成优势子群和普通子群;然后对优势子群继续用改进的粒子群算法进行更新,对普通子群利用差分进化算法进行变异,生成具有高活性的子群;最后利用混合的差分进化算法进行个群更新,并对节点的推荐信任度进行加权计算。仿真结果证明,改进算法具有良好的鲁棒性,同时执行效率高,当节点跳级数较少时,可保证系统的信任度可靠。

In this paper, the problem of network communication nodes optimization was investigated. In P2P network system, frequent exiting of nodes can easily lead to decline in the stability of the network system and cause the trust evaluation of nodes more difficult. To solve the problem, a half coevolution algorithm was introduced to the process of particle swarm evolution for solving of P2P network nodes trust path to make half eoevolution mutant for the individual activity of the population. The optimization algorithm for P2P network node trust pathwas proposed based on half eoevolution. In this algorithm, firstly, we used the improved particle swarm algorithm based on seed spacing and weight coefficient modification to make initial iterative optimization solution for P2P network node trust path, and generated preponderant subgroup and normal subgroup according to average fitness. Then, the improved particle swarm optimization was used to continually update the preponderant subgroup; the differential evolution algorithm was used to make mutant for normal subgroup to generate a high activity subgroup. Finally, the hybrid differential evolu- tion algorithm was applied to update individual group, and the recommended trust of nodes was made by weighted cal- culation. Simulation results show that the improved algorithm has good robustness and high efficiency in the imple- mentation of the algorithm. Furthermore, when the number of node jumping is less, the reliable of the system can be trusted.