Multi-Hop Clustering and Routing Protocol Based on Enhanced Snake Optimizer and Golden Jackal Optimization in WSNs

Zhen Wang, Jin Duan, Pengzhan Xing2024 · DOI

TL;DR

This paper introduces ESO-GJO, a novel energy-efficient multi-hop routing protocol for Wireless Sensor Networks (WSNs). By combining enhanced Snake Optimizer (SO) for cluster head selection and Golden Jackal Optimization (GJO) for routing, ESO-GJO significantly extends network lifetime and reduces energy consumption compared to existing methods.

What Problem Does This Paper Address?

Wireless Sensor Networks (WSNs) are crucial for environmental monitoring and various applications, but their limited battery life and the difficulty of node replacement pose a significant challenge. Energy consumption, particularly during wireless communication, is the primary bottleneck. Existing routing protocols often struggle to balance energy efficiency, network lifetime, and efficient data transmission, especially in large-scale or heterogeneous WSNs. This paper aims to address this by developing a more robust and energy-aware multi-hop clustering and routing protocol.

Methods

The proposed ESO-GJO protocol integrates two meta-heuristic optimization algorithms:

  • Enhanced Snake Optimizer (ESO) for Cluster Head (CH) Selection:
    • The traditional Snake Optimizer (SO) is enhanced by incorporating a Brownian motion function during the exploitation phase. This aims to prevent the algorithm from getting stuck in local optima.
  • The fitness function for CH selection considers multiple parameters: energy consumption of the CH, the node degree of the CH, and the distance between the node and the Base Station (BS). This multi-faceted approach ensures that CHs are chosen not only based on their energy reserves but also their connectivity and proximity to the BS.
  • Golden Jackal Optimization (GJO) for Multi-Hop Routing:
    • After CHs are selected, the Golden Jackal Optimization (GJO) algorithm is employed to determine the optimal multi-hop routing path from the CHs to the BS.
  • To improve population diversity and avoid premature convergence, the GJO algorithm's initial population is seeded using a backward learning technique.
  • The fitness function for GJO considers the distance between CH and BS, the ratio of this distance to the inter-cluster head distance, and the residual energy of the CH. This ensures that the chosen routes are both energy-efficient and minimize transmission distances.

Datasets

Not discussed in the paper. The paper relies on simulated network scenarios for evaluation.

Key Insights and Ideas

  • Synergistic Optimization: Combining two distinct optimization algorithms (ESO for CH selection and GJO for routing) creates a powerful, holistic approach to WSN energy management.
  • Addressing Local Optima: The integration of Brownian motion into the SO algorithm is a clever way to improve exploration and avoid suboptimal CH selections, a common issue in many optimization algorithms.
  • Multi-faceted Fitness Functions: The use of comprehensive fitness functions that consider energy, distance, and node degree for CH selection, and distance and residual energy for routing, is crucial for achieving superior performance.
  • Backward Learning for GJO: Enhancing GJO with backward learning for initial population generation is a key step to improve its exploration capabilities and prevent premature convergence, leading to better routing paths.

Key Results

The ESO-GJO protocol demonstrated superior performance across various metrics compared to PSO-LSA, EPOA-CHS, LEACH-IACA, and LEACH-ANT in simulations with 100, 200, and 300 nodes.

  • Residual Energy: ESO-GJO consistently maintained higher residual energy for surviving nodes across all tested scenarios, indicating significantly better energy conservation. For instance, with 100 nodes, ESO-GJO had 0.08 J remaining after 1500 cycles, while LEACH-IACA and LEACH-ANT died out much earlier.
  • Network Lifetime: The protocol significantly extended network lifetime, measured by metrics like First Node Death (FND), Half Nodes Death (HND), Most Nodes Death (MND), and Last Node Death (LND). For 200 nodes, ESO-GJO extended lifetime by up to 192.2% compared to other protocols.
  • Live Nodes Number: The number of live nodes in the network remained higher for longer periods with ESO-GJO, directly reflecting its improved energy efficiency and network longevity.
  • Packet Delivery: ESO-GJO achieved higher network throughput, meaning more data packets were successfully delivered to the BS. For 300 nodes, it outperformed LEACH-IACA and LEACH-ANT by over 300% and 149%, respectively.
  • Total Energy Consumption: The total energy consumed by the network was lower with ESO-GJO, confirming its efficiency.

Limitations and Future Work

  • FND Performance: The paper acknowledges that the First Node Death (FND) metric, while improved, still falls short in large-scale networks.
  • Future Work: The authors suggest future research should focus on prolonging the FND emergence period for large-scale WSNs, building upon the current findings. They also hint at the need to address network heterogeneity more comprehensively in future enhancements.

Summary updated

BibTeX
@misc{wang2024multi,
  author = {Zhen Wang and Jin Duan and Pengzhan Xing},
  title = {{Multi-Hop Clustering and Routing Protocol Based on Enhanced Snake Optimizer and Golden Jackal Optimization in WSNs}},
  year = {2024},
  doi = {10.3390/s24041348},
}
Figure