论文标题
智能连接的水下传感器网络的耐故障备份聚类算法
Fault-Tolerant Backup Clustering Algorithm for Smart Connected Underwater Sensor Networks
论文作者
论文摘要
本文通过提出能源有效的层次拓扑意识群集路由（EEHTAC）协议来解决水下传感器网络的频繁集群和频繁的簇头（CH）故障问题。在本文中，为EEHTAC操作开发了易耐故障备份聚类（FTBC）算法和多参数群集形成（MPCF）模型。 MPCF模型通过集成多个参数以实现有效的聚类过程来解决群集形成性能差的问题。 FTBC算法解决了频繁的CH失败问题，以避免数据传输的中断。使用普通，高道和高路由开销网络方案评估MPCF模型的性能。用于此分析的性能指标是时间拓扑变异比（TTVR），CH负载分布（CLD）和集群稳定性（STB）。获得的结果表明，以90年代的保留周期运行可实现每轮更好的CH值，并以25.69％，55.56％和60％的ttvr，cld和STB的值改善MPCF工艺。基于FTBC的EEHTAC的性能相对于能量平衡的不等分层聚类（EULC）方案进行了评估。此评估采用的性能指标是路由开销（ω），端到端延迟（δ），恢复的CH失败（CFR），检测到的CH失败（CFD），接收到数据包（θ）和能量消耗（σ）。参照最佳的值，EEHTAC的性能提高了58.40％，29.94％，81.33％，28.02％，86.65％和54.35％的性能在ω，δ，δ，Δ，cfr，cfd，θ和σ方面比EULC变体的变化比EULC变体的改进分别提高了54.35％。获得的结果显示，MPCF模型对于集群形成性能有效，基于FTBC的EEHTAC协议可以在现有的CBR协议上有效地性能。
This paper addresses poor cluster formation and frequent Cluster Head (CH) failure issues of underwater sensor networks by proposing an energy-efficient hierarchical topology-aware clustering routing (EEHTAC) protocol. In this paper, fault-tolerant backup clustering (FTBC) algorithms and multi-parameter cluster formation (MPCF) model were developed for the EEHTAC operation. The MPCF model tackles the issue of poor cluster formation performance by integrating multiple parameters to achieve effective clustering process. The FTBC algorithms tackle the issue of frequent CH failures to avoid interruption in data transmission. Performance of the MPCF model was evaluated using normal, high-fault, and high routing overhead network scenarios. Performance metrics employed for this analysis are temporal topology variation ratio (TTVR), CH load distribution (CLD), and cluster stability (STB). Obtained results show that operating with a CH retention period of 90s achieves better CH duty cycling per round and improves the MPCF process with values of 25.69%, 55.56%, and 60% for TTVR, CLD, and STB respectively. Performance of the FTBC-based EEHTAC was evaluated relative to Energy-balanced Unequal Layering Clustering (EULC) protocol. Performance indicators adopted for this evaluation are routing overhead (Ω), end to end delay (Δ), CH failures recovered (CFR), CH failures detected (CFD), received packets (θ), and energy consumption (Σ). With reference to the best obtained values, EEHTAC demonstrated performance improvement of 58.40%, 29.94%, 81.33%, 28.02%, 86.65%, and 54.35% over EULC variants in terms of Ω, Δ, CFR, CFD, θ, and Σ respectively. Obtained results displayed that the MPCF model is efficient for cluster formation performance and the FTBC-based EEHTAC protocol can perform effectively well against an existing CBR protocol.
