Abstract

The topology management classifiers consist of several methods, such as the typical clustering-based method excelled in wireless network partitioning. However, most algorithms appear load unbalanced in the application of irregular network, resulting “energy hot zone” phenomenon. This paper proposes an improved intelligent clustering algorithm and applies it to the complex water system environment. Firstly, we build a new energy consumption model for wireless transmission network, and design a genetic clustering strategy via the minimum energy consumption principle. Secondly, we introduce the P matrix coding approach considering the search scale, so as to avoid the squared increasing relationship between the searching space and the data calculation. Thirdly, we employ adaptive genetic operator to enhance the directivity of the searching space, and utilize a fuzzy modified operator to enhance the accuracy of the cluster head selection, which may ensure the iterative efficiency. Through numerical simulations, empirical results show better performance than traditional methods in load balancing and clustering efficiency, which can effectively improve the network convergence speed and extend the network lifetime.

中文翻译：

---

一种改进的不规则无线网络智能聚类算法

摘要

拓扑管理分类器由多种方法组成，例如在无线网络分区方面表现出色的典型的基于聚类的方法。但是，大多数算法在不规则网络的应用中出现负载不平衡的情况，从而导致“能量热区”现象。提出了一种改进的智能聚类算法，并将其应用于复杂的水系统环境。首先，我们建立了一种新的无线传输网络能耗模型，并根据最小能耗原理设计了遗传聚类策略。其次，我们介绍P考虑搜索尺度的矩阵编码方法，避免了搜索空间与数据计算之间平方增加的关系。第三，采用自适应遗传算子来增强搜索空间的方向性，并利用模糊修正算子来提高簇头选择的准确性，从而可以保证迭代效率。通过数值模拟，实验结果表明，在负载均衡和集群效率方面，性能优于传统方法，可以有效提高网络收敛速度，延长网络寿命。