针对分布式无线传感器网络能量消耗和数据通信实时性问题,利用无线传感器网络的特点,提出了新型的蚁群路由算法,即基于Energy*Delay模型的蚁群算法(E8LDANTS),算法中人工蚂蚁通过在线延迟的方式进行数据交换,并收集邻居节点状态和网络路由信息,以此建立起最佳路由表,使得在每次传输固定大小的数据和消耗相同能量的情况下数据包传输时延最小,由于在无线通信系统中,能量消耗和传输时延是两个相互对立特征量,在此运用加强学习(RI。)的方法来训练该模型,通过仿真试验与传统的LEACH和PEGASIS路由算法比较,结果表明,该算法的有效性比LEACH算法提高12倍,也优于PEGASIS算法3倍左右。
A new ant routing algorithm based on Energy * Delay model (E&D ANTS) was proposed utilizing the characteristic of wireless sensor network to solve the problems of energy consumption and real-time data transmission in distributed wireless sensor network. Artificial ants exchanged data, collected neighbor node statement and network routing information in the mode of online delay, then the best routing table was constructed to minimize the time delay in transferring a fixed number of data in an energy-constrained manner in one round. The reinforcement learning (RL) algorithm was employed to train the model for the energy consumption and the transmission delay are two relative independent characteristic quantities in wireless network system. Simulation results show that the algorithm performs twelve times better than low-energy adaptive clustering hierarchy (LEACH) and three times better than power-efficient gathering in sensor information system (PEGASIS) in terms of energy cost and delay per round.