Abstract
Solar-powered nodes employed in farmland wireless sensor networks offer significant energy advantages but come with high implementation costs. To balance cost and performance, networks often employ a mix of solar-powered and battery-powered nodes. This paper presents an improved clustering algorithm based on the LEACH protocol designed for such heterogeneous environments. The algorithm classifies nodes into three types based on their energy characteristics and prioritizes solar-powered nodes for the energy-intensive role of cluster head through a two-stage election process. Simulation results demonstrate that this approach improves solar energy utilization, reduces node mortality, and significantly extends the overall network lifecycle.
Introduction
Monitoring large-scale farmland requires wireless sensor networks (WSNs) capable of long-term operation in remote areas. Traditional battery-powered networks are limited by finite energy supplies, leading to "coverage holes" as nodes fail. While solar-powered nodes provide a renewable alternative, their high cost and dependence on weather conditions make a purely solar network impractical. This research addresses the challenge of energy heterogeneity: how to effectively coordinate a hybrid network of solar and battery nodes to maximize lifespan while minimizing costs.
Proposed Clustering Mechanism
The proposed algorithm enhances the Low Energy Adaptive Clustering Hierarchy (LEACH) protocol. While standard LEACH selects cluster heads (CHs) randomly, this improved version uses a "solar-first" priority system to ensure that nodes with renewable energy carry the heaviest data-routing burdens.
Node Classification and Election
At the beginning of each round, the sink node categorizes nodes based on their energy source and current levels:
- Quasi-cluster head solar nodes: Solar nodes with energy above a specific threshold ($E_{sthr}$).
- Downgrade solar nodes: Solar nodes with insufficient energy (e.g., during nighttime or rainy weather).
- Battery nodes: Nodes with a non-renewable, decreasing energy supply.
The election process occurs in two stages. First, the algorithm attempts to fill the required number of CH positions using only the "Quasi-cluster" solar nodes. If the number of available solar nodes is insufficient to meet the network's needs, a second election is held among the battery-powered nodes to fill the remaining slots.
Simulation and Performance Analysis
The algorithm was tested via MATLAB simulations in a 100m × 100m field with 100 nodes (20% solar, 80% battery). The simulation included environmental modeling of day/night cycles and a multi-round "rainy" period where solar charging was unavailable, forcing solar nodes to rely on their internal reserves.
Key Results
The improved algorithm was compared against the standard LEACH protocol over 5,000 rounds.
- Energy Efficiency: The hybrid approach successfully offloaded the most energy-intensive tasks to solar nodes during sunny periods, preserving the limited reserves of battery nodes for when they are most needed.
- Residual Energy: As shown in the simulation results, the improved algorithm maintained a significantly higher average residual energy across the network compared to LEACH throughout the entire lifecycle.
- Network Lifetime: By prioritizing renewable energy, the algorithm delayed the "first node death" and kept approximately 40% of the network operational by round 5000, whereas the standard LEACH network had almost entirely failed by that point.
Conclusion
By intelligently managing energy heterogeneity, the proposed clustering mechanism offers a robust solution for farmland monitoring. Prioritizing solar-powered nodes for high-energy roles ensures that battery resources are conserved for periods of low solar availability. Future research will focus on optimizing node placement and adapting the algorithm for mobile sensor nodes in smart agriculture applications.
Summary updated