Scholarlink

<p>A new decentralized data collection technique is proposed for multi-sink wireless sensor networks (WSNs) with the objective of energy and delay reduction. The K-means algorithm which is based on the position of sensors partitions the network into k clusters. The sensor that is the closest to the centroid is voted as the cluster head (CH) in each cluster. Two mobile data collectors (MDCs) are assigned to each other to disseminate the data, one MDC gets the CHs as the other one does. The MDCs determine the route from their designated CHs with the help of a nearest neighbour heuristic and then each tour is improved with 2-opt local optimization (a simple TSP improvement) to minimize the total distance. NS-3 simulations allowed us to set up 100 static sensors (in a 100 × 100&#xa0;m area) and two MDCs (with constant velocity mobility). Compared to the single-sink baselines, our experiments indicate that the two-sink approach with optimized tours has a pronounced effect in terms of reducing the end-to-end delay and total energy consumption. The average data collection delay, network energy expenditure, and MDC tour lengths are reported, showcasing the effectiveness of centroid-based clustering and 2-opt optimization in load balancing and network lifetime prolongation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Hybrid K-Means and 2-opt Metaheuristic for Trajectory Optimization in Dual-Sink WSNs

  • 期刊论文

  • 发表日期：

    2026年5月27日

查看全文
  • Satish Kumar Maurya,
  • Vinod Jain,
  • Jagdish Chandra Patni,
  • Vivek Kumar Prasad,
  • Om Prakash Pal

摘要

A new decentralized data collection technique is proposed for multi-sink wireless sensor networks (WSNs) with the objective of energy and delay reduction. The K-means algorithm which is based on the position of sensors partitions the network into k clusters. The sensor that is the closest to the centroid is voted as the cluster head (CH) in each cluster. Two mobile data collectors (MDCs) are assigned to each other to disseminate the data, one MDC gets the CHs as the other one does. The MDCs determine the route from their designated CHs with the help of a nearest neighbour heuristic and then each tour is improved with 2-opt local optimization (a simple TSP improvement) to minimize the total distance. NS-3 simulations allowed us to set up 100 static sensors (in a 100 × 100 m area) and two MDCs (with constant velocity mobility). Compared to the single-sink baselines, our experiments indicate that the two-sink approach with optimized tours has a pronounced effect in terms of reducing the end-to-end delay and total energy consumption. The average data collection delay, network energy expenditure, and MDC tour lengths are reported, showcasing the effectiveness of centroid-based clustering and 2-opt optimization in load balancing and network lifetime prolongation.