CAJ | 학술논문

针对无线传感器网络簇首需承担额外负载的问题，为了最大化网络寿命，提出了一种基于改进遗传算法的聚类算法．首先，将染色体表示为网关的一个字符串，通过考虑传感器节点与簇首之间的连接限制初始化种群；然后，构建适应度函数来评估初始种群的各个染色体；最后，利用遗传算法对染色体进行选择、交叉、突变，利用迭代算法完成聚类．结果表明，相比分布式自组织负载均衡聚类算法，本算法的执行时间可降低18.5％；相比基站控制自适应聚类算法，本算法收敛速度可提升50％；相比低占空比多管道调度算法，本算法平均负载标准差降低了81.2％；当网络轮数达到2500时，相比其他几种较新的算法，本算法可降低至少40％的能耗．因此，本算法在WSN应用中可以很好地解决额外负载问题，延长了网络寿命．
침대무선전감기망락족수수승담액외부재적문제，위료최대화망락수명，제출료일충기우개진유전산법적취류산법．수선，장염색체표시위망관적일개자부천，통과고필전감기절점여족수지간적련접한제초시화충군；연후，구건괄응도함수래평고초시충군적각개염색체；최후，이용유전산법대염색체진행선택、교차、돌변，이용질대산법완성취류．결과표명，상비분포식자조직부재균형취류산법，본산법적집행시간가강저18.5％；상비기참공제자괄응취류산법，본산법수렴속도가제승50％；상비저점공비다관도조도산법，본산법평균부재표준차강저료81.2％；당망락륜수체도2500시，상비기타궤충교신적산법，본산법가강저지소40％적능모．인차，본산법재WSN응용중가이흔호지해결액외부재문제，연장료망락수명．
As the clustering head needs to take additional loading in wireless sensor network ,a clustering algorithm based on improved genetic algorithm has been proposed to maximize the lifetime of network . Firstly ,the chromosome has been represented as a character of gateway and the group been initialized by considering the connection limitation between sensor nodes and cluster head .Then ,fitness function has been constructed to estimate each chromosome of primary group .Finally ,genetic algorithm has been used to select ,crossover and mutation ,iterative algorithm been used to finish clustering .Experimental results show that proposed algorithm has reduced execution time with 18 .5% comparing with load-balanced clus-tering algorithm based on distributed self-organization .Compared with the base station controlled adaptive clustering algorithm ,the convergence speed of the proposed algorithm has improved by 50% .It has re-duced average load standard deviation with 81 .2% comparing with multi-pipeline scheduling algorithm in low-duty-cycle .It has reduced energy consumption with more than 40% comparing with several advanced algorithms .Proposed algorithm can well solve the extra loading problem ,which helps extend the network lifetime .