Energy Saving Routing Algorithm for Wireless Sensor Networks Based on Minimum Spanning Hyper Tree

With the rapid development of wireless sensor networks (WSNs), designing energy-efficient routing protocols has become essential to prolong network lifetime. This paper proposes a minimum spanning tree-based energy-saving routing algorithm for WSNs. First, sensor nodes are clustered using the LEACH protocol and minimum spanning trees are constructed within clusters and between cluster heads. The spanning tree edge weights are optimized considering transmission energy, residual energy, and energy consumption rate. This avoids channel competition and improves transmission efficiency. An energy-saving routing model is then built whereby deep rein-forcement learning (DRL) agents calculate paths optimizing the energy utilization rate. The DRL reward function integrates network performance metrics like energy consumption, delay, and packet loss. Experiments show the proposed approach leads to 10-15W lower average switch energy consumption compared to existing methods. The throughput is high since overloaded shortest paths are avoided. The average path length is close to shortest path algorithms while maintaining energy efficiency. In summary, the proposed minimum spanning tree-based routing algorithm successfully achieves energy-saving goals for WSNs while guaranteeing network performance. It provides an efficient and adaptive routing solution for resource-constrained WSNs.

Automated summary

This paper proposes a minimum spanning tree-based energy-saving routing algorithm for wireless sensor networks. It clusters sensor nodes and constructs minimum spanning trees to optimize energy efficiency. Using deep reinforcement learning, it calculates paths that optimize energy utilization rate by considering energy consumption, delay, and packet loss. The proposed algorithm successfully achieves energy-saving goals for WSNs while guaranteeing network performance.