Clustering Nodes Selection Mechanism in Energy Heterogeneous Farmland Wireless Sensor Networks

Dayong Liu, Chunjiang Zhao, Huarui Wu, Feifei Li2016 · DOI

Abstract

Solar-powered nodes are increasingly integrated into farmland wireless sensor networks (WSNs) to extend operational lifetimes. However, the intermittent and nonlinear nature of solar energy acquisition complicates network management. This paper presents a fuzzy-logic-based selection mechanism for cluster heads in energy-heterogeneous WSNs. By predicting solar energy availability from weather forecasts and incorporating multiple node factors, the proposed EARA protocol optimizes cluster formation. Simulation results demonstrate that this approach significantly enhances network throughput and energy efficiency compared to traditional protocols.

Introduction

Wireless Sensor Networks (WSNs) in agricultural settings often face the challenge of remote deployment where battery replacement is impractical. Energy-heterogeneous networks, combining standard battery-powered nodes with solar-harvesting nodes, offer a solution. Yet, traditional routing protocols like LEACH (Low-Energy Adaptive Clustering Hierarchy) often fail to account for the volatility of solar power. A solar node might be highly capable during sunny periods but become a bottleneck during extended overcast weather. To address this, a mechanism is needed that dynamically adjusts to environmental conditions by translating qualitative weather data into actionable routing logic.

Methodology: The EARA Protocol

The proposed EARA protocol introduces a sophisticated selection process for cluster heads based on two primary innovations:

  1. Fuzzy Solar Prediction: The system translates qualitative weather forecasts (e.g., "sunny," "cloudy," "overcast") into quantitative energy predictions. Using fuzzy set theory, these descriptions are mapped to a "Charge Coefficient" ($B_j$), which models the expected solar intake based on light occlusion.
  2. Multi-Factor Cluster Selection: Cluster heads are not chosen by battery level alone. Instead, the protocol evaluates nodes based on four weighted parameters: remaining energy, node type (solar vs. battery), node degree (neighbor count), and the predicted charge coefficient. The algorithm employs a "maximum-minimum synthesis" method to calculate final weights and organize nodes into optimal clusters.

Experimental Setup

The performance of the EARA protocol was evaluated via Matlab simulations. The environment modeled a 100m x 100m farmland area with 100 randomly deployed nodes (20 solar-powered and 80 battery-powered). Each node began with an initial energy of 2 Joules. Hardware parameters included a 1000-bit data packet size and specific power consumption rates for transmission (50nJ), reception (10nJ), and idle states (0.013pJ) to model real-world sensor behavior.

Results and Discussion

The EARA protocol was compared against the standard LEACH algorithm across two key metrics: throughput and energy longevity.

Network Throughput

The EARA protocol maintained a consistently higher throughput than LEACH. While LEACH's data transmission rate began to plateau as nodes depleted their energy, EARA continued to facilitate successful bit transmissions, effectively doubling the total throughput by the 3,000-round mark. This suggests that the fuzzy selection mechanism successfully identifies the most capable nodes to handle heavy traffic.

Network throughput comparison

Energy Efficiency

The network's residual energy ratio also favored the proposed method. By intelligently offloading heavy tasks to solar nodes only when recharge was predicted, EARA preserved more of the network's total energy. By round 5,000, the EARA-managed network retained a significantly higher percentage of its initial energy compared to the LEACH-managed network, effectively extending the operational life of the monitoring system.

Network residual energy ratio

Conclusion

By bridging the gap between unpredictable weather patterns and network routing logic, the EARA protocol offers a robust solution for agricultural WSNs. The use of fuzzy logic to interpret weather forecasts allows the network to proactively manage its energy resources, leading to superior throughput and extended operational life. Future work will focus on refining weather granularity and testing the system in real-world large-scale deployments across diverse environmental conditions.

Summary updated

Abstract

Solar-powered nodes have been employed in farmland wireless sensor networks to prolong the lifetime, which can increase the cost at the same time. The energy acquisition mode of solar-powered node is intermittent and nonlinear, because it is easy to be influenced by the external factors. In this paper, a fuzzy prediction method of the solar energy using the weather forecast is proposed, and then a fuzzy selection method to distinguish the different types of nodes is proposed. Simulation results show that the proposed method can improve the network throughput and reduce the energy consumption of nodes.
BibTeX
@misc{liu2016clustering,
  author = {Dayong Liu and Chunjiang Zhao and Huarui Wu and Feifei Li},
  title = {{Clustering Nodes Selection Mechanism in Energy Heterogeneous Farmland Wireless Sensor Networks}},
  year = {2016},
  doi = {10.2991/icmemtc-16.2016.224},
}
Figure