摘要
为了解决无线传感器网络节点分布不均,导致有效网络覆盖率较低的问题,提出一种融合莱维飞行与混合变异的蝠鲼觅食优化传感器节点覆盖策略M⁃MRFO。首先,在蝠鲼种群初始化生成方面引入广义对立学习机制,提高种群在搜索空间内的多样性和算法遍历性;其次,结合莱维(Levy)飞行机制对算法的权重因子和翻滚因子进行调整,通过Levy飞行的随机跳跃式搜索提高种群的全局寻优能力;最后,提出针对精英个体的高斯分布和柯西分布混合变异方法,使算法具备跳离局部最优的能力。将改进算法应用于传感器节点的网络覆盖优化中,利用蝠鲼种群启发式觅食行为模式对节点部署位置迭代寻优。实验结果表明,与标准蝠鲼觅食优化算法MRFO、改进差分进化算法IDEA和混合改进蚁狮算法MS⁃ALO相比,改进算法M⁃MRFO能够有效降低节点冗余,更均匀地实现节点部署,提高网络覆盖率。
In order to improve the node coverage of wireless sensor networks,a sensor node coverage optimization method based on im⁃proved manta ray foraging optimization algorithm with Levy flight and hybrid mutation is proposed.Firstly,the generalized opposition⁃learning mechanism is applied to initialize the manta ray population to improve the diversity of the population in search space and the ergodicity of the algorithm.Secondly,combined with Levy flight mechanism,the weight factor and roll factor of the algorithm are adjus⁃ted,and the global optimization ability of the population is improved through the random jump search of Levy flight.Then,a hybrid mu⁃tation method of Gaussian distribution and Cauchy distribution for elite individuals is proposed to make the algorithm have the ability to jump away from local optimum.The improved algorithm is applied to solve the network coverage optimization of sensor nodes,and the node location is iteratively optimized by adopting the foraging behavior mode of manta ray population.The results show that compared with the standard manta ray foraging optimization algorithm(MRFO),the hybrid improved ant lion algorithm of MS⁃ALO and the im⁃proved differential evolution algorithm(IDEA),the improved algorithm of M⁃MRFO can efficiently reduce nodes redundancy,better real⁃ize the uniform distribution of nodes,and improve the network coverage.
作者
许杰
汤显峰
XU Jie;TANG Xianfeng(The Second Affiliated Hospital,School of Medicine,Zhejiang University,Hangzhou Zhejiang 310009,China;Center of Information Technology,Zhejiang University,Hangzhou Zhejiang 310027,China)
出处
《传感技术学报》
CAS
CSCD
北大核心
2023年第4期635-645,共11页
Chinese Journal of Sensors and Actuators
基金
国家自然科学基金项目(61602141,61503336)。
关键词
无线传感器网络
节点覆盖
蝠鲼觅食优化算法
对立学习
莱维飞行
高斯分布
wireless sensor network
nodes coverage
manta ray foraging optimization algorithm
opposition learning
Levy flight
Gauss distribution
