基于蚁群算法和BP神经网络的信道分配策略的研究

发布时间：2018-04-03 20:14

本文选题：无线传感器网络　切入点：信道分配机制　出处：《传感技术学报》2016年03期

【摘要】：研究无线传感器网络信道分配策略的主要目标是提高网络吞吐量和容量,减小网络的传输时延,最大限度的利用有限的网络带宽资源。多信道MAC协议的应用,可以有效地提高网络通信的可靠性和吞吐量,以及解决由于信道受干扰而造成的网络瘫痪等问题。根据无线传感器网络多信道的特点提出了一种基于蚁群算法的动态反馈负载均衡信道分配策略。本策略首先应用BP神经网络对信道负载情况进行预测,然后通过基于蚁群算法的负载均衡算法对信道进行筛选,最后利用最大离散化算法进行信道分配。在NS2平台下对所设计的协议进行了仿真实现,并与应用最为广泛的多信道MMAC协议以及SMAC进行了对比分析。根据仿真结果可知,本文设计的MAC协议在网络吞吐量、网络传输时延等性能方面比MMAC协议及SMAC都有了很大程度的提升。可以有效减小网络传输时延,提高网络吞吐量和抗干扰能力。
[Abstract]:The main goal of the channel allocation strategy in wireless sensor networks is to improve the throughput and capacity of the network, reduce the transmission delay of the network, and maximize the use of limited network bandwidth resources.The application of multi-channel MAC protocol can effectively improve the reliability and throughput of network communication and solve the network paralysis caused by channel interference.According to the multi-channel characteristics of wireless sensor networks, a dynamic feedback load balancing channel allocation strategy based on ant colony algorithm is proposed.Firstly, BP neural network is used to predict the channel load, then the load balancing algorithm based on ant colony algorithm is used to filter the channel, and the maximum discretization algorithm is used to allocate the channel.The designed protocol is simulated on NS2 platform and compared with the most widely used multi-channel MMAC protocol and SMAC.According to the simulation results, the performance of the proposed MAC protocol is greatly improved than that of the MMAC protocol and the SMAC protocol in terms of network throughput, network transmission delay and so on.It can effectively reduce network transmission delay and improve network throughput and anti-interference ability.
【作者单位】：华北电力大学(保定)控制与计算机工程学院;
【基金】：国家自然科学基金项目(60974125)
【分类号】：TP212.9;TN929.5;TP18

【相似文献】