Optimal LEACH Protocol with Improved Bat Algorithm in Wireless Sensor Networks

Xingjuan Cai, Youqiang Sun, Zhihua Cui, Wensheng Zhang, Jinjun ChenKSII Trans. Internet Inf. Syst 2019 · OpenReview

Abstract

Wireless Sensor Networks (WSNs) are critical for various monitoring applications but are severely limited by node battery life. The Low-Energy Adaptive Clustering Hierarchy (LEACH) protocol is a foundational approach to energy conservation, yet its random cluster head selection often leads to inefficient energy distribution. This paper introduces FTBA-TC-LEACH, a protocol that optimizes cluster head selection using an improved Bat Algorithm. By integrating a "triangle-flipping" strategy for global search and a "curve decline strategy" for parameter tuning, the proposed method significantly extends network lifetime and improves energy efficiency compared to traditional metaheuristic approaches.

Introduction

Wireless Sensor Networks (WSNs) consist of numerous low-power nodes that collect and transmit data to a base station. The primary challenge in WSN design is maximizing network longevity, as nodes typically operate on non-rechargeable batteries. The LEACH protocol addresses this by organizing nodes into clusters, where a Cluster Head (CH) aggregates data from its members and transmits it to the base station. However, the stochastic nature of CH selection in standard LEACH can lead to the premature death of nodes with low energy or those located far from the base station.

The FTBA-TC-LEACH Protocol

To overcome the limitations of random selection, the authors propose FTBA-TC-LEACH. This protocol operates in cycles, dividing time into setup and communication phases to ensure organized data transmission.

Operational cycle structure of the proposed protocol, illustrating the division of time into setup (Head) and data transmission (Communication) phases.

The core of the protocol is an improved Bat Algorithm (BA), which mimics the echolocation of bats to find optimal CH configurations. The improvements are two-fold:

  1. Fast Bat Algorithm (FTBA): A "triangle-flipping strategy" is employed to enhance global search capabilities, preventing the algorithm from converging on local optima.
  2. Curve Strategy (TC): A non-linear decline strategy for the search radius parameter is introduced to balance exploration and exploitation.

Curve Strategy for Parameter Optimization

The effectiveness of the Bat Algorithm depends heavily on how search parameters evolve over generations. The authors investigated various mathematical curves for the decline of the search parameter $\tau$.

Comparison of different curve decline strategies (linear, convex, and concave) for the search parameter tau across generations.

The research found that a concave curve (where the parameter decreases slowly initially and accelerates later) provides the best results. This allows the "bats" to explore a wide area in the early stages of the search and converge precisely on the optimal solution in the final rounds.

Fitness Function and Selection Criteria

The algorithm evaluates potential Cluster Heads using a multi-objective fitness function. The selection is based on three primary factors:

  • Intra-cluster Distance: Minimizing the distance between member nodes and their CH.
  • Base Station Distance: Minimizing the distance from the CH to the sink.
  • Residual Energy: Prioritizing nodes with higher-than-average remaining battery power to prevent "weak" nodes from being overburdened.

Experimental Results

The protocol was validated through numerical optimization benchmarks and WSN simulations.

Numerical Benchmarks

Using the CEC 2013 dataset, the FTBA-TC algorithm was tested against standard BA, Particle Swarm Optimization (PSO), and Cuckoo Search (CS). FTBA-TC consistently outperformed these metaheuristics across 28 complex mathematical functions, demonstrating superior robustness and convergence speed.

WSN Simulation

In a MATLAB simulation of 100 nodes, FTBA-TC-LEACH showed dramatic improvements over the baseline:

  • Node Survival: After 2,000 rounds, FTBA-TC-LEACH maintained 11 surviving nodes, compared to only 6 for standard LEACH and 8 for the standard Bat Algorithm (SBA-LEACH).
  • Energy Efficiency: The total residual energy of the network using FTBA-TC-LEACH was 1.63 times higher than that of standard LEACH.

Conclusion

The FTBA-TC-LEACH protocol represents a significant advancement in WSN energy management. By replacing random CH selection with a sophisticated, curve-optimized Bat Algorithm, the researchers achieved a more balanced energy distribution and extended network longevity. Future work will explore the application of this algorithm to mobile node environments and node localization in GPS-denied areas.

Summary updated

BibTeX
@article{cai2019optimal,
  author = {Xingjuan Cai and Youqiang Sun and Zhihua Cui and Wensheng Zhang and Jinjun Chen},
  title = {{Optimal LEACH Protocol with Improved Bat Algorithm in Wireless Sensor Networks}},
  journal = {KSII Trans. Internet Inf. Syst.},
  year = {2019},
  url = {https://openreview.net/forum?id=jZTdZHsguA},
}
Figure