A Study on the Clustering Technology of Underwater Isomorphic Sensor Networks Based on Energy Balance

Fei Wang, Liming Wang, Yan Han, Bin Liu, Jian Wang, Xinyan Su2014 · DOI

TL;DR

This paper introduces a novel clustering approach for underwater sensor networks (UWSNs) called Underwater Isomorphic Sensor Networks (UWISNs). By dynamically electing cluster heads based on energy balance and proximity, UWISNs aim to significantly extend network lifetime and improve energy efficiency compared to traditional heterogeneous UWSNs.

What Problem Does This Paper Address?

Underwater sensor networks (UWSNs) are crucial for applications like military surveillance, resource detection, and environmental monitoring. A major challenge in UWSNs is the limited power of sensor nodes, necessitating energy-efficient communication strategies to prolong network lifetime. Traditional approaches often rely on Underwater Heterogeneous Sensor Networks (UWHSNs), where specific high-energy nodes act as cluster heads. However, this method has drawbacks: the failure of a cluster head can disrupt the network, it requires a large number of high-energy nodes for extensive coverage (increasing cost), and node mobility due to ocean currents can lead to inefficient communication and increased energy consumption as nodes drift away from their designated cluster heads. This paper addresses these limitations by proposing a more flexible and robust clustering mechanism.

Methods

The paper proposes an Underwater Isomorphic Sensor Network (UWISN) clustering technology. Unlike UWHSNs, UWISNs treat all sensor nodes as isomorphic, meaning they have the same initial energy and structure. The core of the method involves:

  1. Optimal Cluster Head Election: Instead of fixed cluster heads, UWISNs periodically elect the "best" cluster heads based on the current network environment. This election prioritizes nodes with higher residual energy and shorter distances to a theoretical central point.
  2. Fuzzy C-Means (FCM) Algorithm: The Fuzzy C-Means algorithm is employed to determine the optimal number of clusters and the initial locations of theoretical cluster heads. FCM is used to minimize the weighted distance between cluster members and cluster heads.
  3. Real Cluster Head Election: After theoretical cluster heads are identified, real cluster heads are elected from the sensor nodes within the vicinity. This election uses a formula that balances the distance of a node to the theoretical cluster head and its remaining energy, aiming for energy efficiency.
  4. Energy Balance Principle: Nodes join clusters based on the principle of energy balance and proximity to elected cluster heads. This ensures that communication is as energy-efficient as possible.
  5. Dynamic Re-election: If a cluster head dies, other nodes in the cluster can take over, and if nodes move due to ocean currents, they can dynamically join new clusters, enhancing network mobility and robustness.

Datasets

Not discussed in the paper. The paper relies on simulations using MATLAB.

Key Insights and Ideas

  • Isomorphic Nodes for Flexibility: Treating all nodes as initially identical simplifies the network architecture and allows any node to potentially become a cluster head, leading to more dynamic and resilient network operation.
  • Energy Balance as a Primary Driver: The core innovation lies in prioritizing energy balance during cluster head election and cluster formation. This directly addresses the limited power of underwater sensors.
  • Dynamic Cluster Head Election: The periodic re-election of cluster heads, rather than relying on static ones, significantly improves network lifetime by distributing the role and preventing premature energy depletion of specific nodes.
  • FCM for Optimal Clustering: The use of Fuzzy C-Means provides a systematic way to determine the optimal number of clusters and initial cluster head locations, laying a solid foundation for the subsequent real cluster head election.
  • Bridging Theoretical and Real Cluster Heads: The paper acknowledges that theoretical optimal locations might not always be ideal in a dynamic environment. The election of "real" cluster heads from actual sensor nodes addresses this practical challenge.

Key Results

The paper presents simulation results demonstrating the effectiveness of the proposed UWISN clustering technology. Key findings include:

  • Improved Energy Efficiency: Simulations show that UWISNs can improve the energy consumption of an underwater sensor network. Figure 6 illustrates the relationship between the number of clusters and energy consumption, indicating an optimal point (around c=5 in their simulations) for minimum energy usage.
  • Comparison of Theoretical vs. Real Cluster Heads: Figure 7 compares the energy consumption of theoretical cluster heads versus real cluster heads. The results show that the energy consumption with real cluster heads is nearly the same as with theoretical ones, validating the election method for real cluster heads.
  • Robustness to Node Distribution: The simulations suggest that the approach is effective even with varying node distributions, conforming to a Poisson distribution.

Limitations and Future Work

  • Simulation-Based: The results are based on simulations using MATLAB, and real-world deployment might reveal different challenges.
  • Optimal Cluster Number: While the paper discusses finding the optimal number of clusters, the exact method for determining this number in all scenarios might require further refinement.
  • Acoustic Signal Attenuation: The paper acknowledges that acoustic signals are attenuated with distance and frequency, and while it incorporates this into energy consumption models, the complexity of underwater acoustic environments could present further challenges.
  • Future Work: The authors suggest that future work could involve exploring more advanced optimization algorithms for cluster head election and investigating the impact of different underwater environments on network performance.

Summary updated

BibTeX
@misc{wang2014study,
  author = {Fei Wang and Liming Wang and Yan Han and Bin Liu and Jian Wang and Xinyan Su},
  title = {{A Study on the Clustering Technology of Underwater Isomorphic Sensor Networks Based on Energy Balance}},
  year = {2014},
  doi = {10.3390/s140712523},
}
Figure