摘要
针对传统OLSR路由算法存在的不足,提出了一种节点剩余能量组合预测的OLSR路由算法(MOLSR)。首先采用回归移动平均模型对节点的剩余能量线性变化特点进行预测,然后采用神经网络对残差的时间序列建立非线性预测模型,对节点的剩余能量非线性变化特点进行预测,最后将两者的预测结果进行相加,并用于进行OLSR路由算法的路由选择中。仿真结果表明,MOLSR路由算法不仅减小了网络开销,有效防止节点剩余能量过早耗完,而且延长了网络的生存时间,具有一定的实际应用价值。
Aiming at the shortcomings of the traditional OLSR routing algorithm,this paper modified OLSR routing protocol( MOLSR) is proposed based on based on residual energy predicted by hybrid model. Firstly,autoregressive moving average model is used to predict residual energy of nodes,secondly,BP neural network is used to establish non linear prediction model of residual time series and describe nonlinear changes features of residual energy nodes,finally the two prediction results are added to get the prediction results and application for the OLSR routing algorithm. The simulation results show that the proposed routing algorithm not only reduces energy consumption and effectively prevent the premature and prolongs the network survival time,has certain practical application value.
出处
《激光杂志》
北大核心
2015年第3期100-103,共4页
Laser Journal
基金
河南省基础与前沿技术研究项目(142300410339)
关键词
移动自组网络
优化链路状态路由算法
回归移动平均
神经网络
mobile ad hoc network
optimized link state routing algorithm
autoregressive integrated moving average
neural network
