Abstract Energy scarcity is one of the most critical problems that occur in wireless sensor networks compared to traditional networks. However, the problem has been partly solved by building the power consumption model of WSN. This paper is based on a simple wireless sensor network model, it gives the optimal location of the CH nodes in a cluster and the best parameter about how to divided the cluster amongst hierarchy networks. As shown last, the proposed scheme can save up to 95% of power consumption.
1. Introduction
Wireless Sensor Networks (WSNs) are typically composed of battery-powered nodes where energy efficiency is paramount due to the difficulty of battery replacement. While clustering techniques—where nodes transmit to a local Cluster Head (CH)—are standard, many protocols like LEACH employ random CH selection. This paper addresses two primary inefficiencies: suboptimal CH placement within a cluster and the lack of mathematical rigor in determining optimal cluster sizes for large-scale networks.
2. Methodology
The study utilizes a mathematical modeling approach based on the power consumption formula $E = d^\sigma$, specifically using $\alpha=2$ (free space model). The network is represented as an $N \times N$ grid where nodes communicate with immediate neighbors.
As illustrated in the modeling diagrams, the researchers analyzed both odd and even grid dimensions to calculate cumulative transmission costs. The methodology involves:
- Geometric Cost Analysis: Calculating the total power required for all nodes in a cluster to reach a CH at specific coordinates $(X, Y)$.
- Hierarchical Optimization: Defining a total cost function $C_{total}(N, x)$ for a two-level hierarchy and utilizing calculus to identify the optimal sub-cluster size $x$.
- Simulation: Validating the mathematical proofs through synthetic node deployment simulations rather than external datasets.
3. Key Insights and Results
The mathematical derivation leads to several fundamental principles for energy-efficient WSN design:
3.1 The "Center of Mass" Rule
The most efficient location for a Cluster Head is the geometric center of the cluster. Moving a CH from the center to the boundary in a 1000-node cluster can double power consumption. Optimizing CH placement alone can save up to 50% of energy.
3.2 The Square Root Scaling Law
For a network of $N$ nodes, the optimal sub-cluster size ($x$) is approximately $\sqrt{N}$. This provides a clear rule of thumb for network engineers: a 10,000-node network should ideally be divided into sub-clusters of approximately 100 nodes each.
3.3 Comparative Efficiency at Scale
The efficiency gap between flat and hierarchical networks grows exponentially with $N$. The ratio of hierarchical to non-hierarchical cost is approximately $2\sqrt{N} / (N+1)$. Empirical results show that at $N = 2000$ nodes, the optimized hierarchical structure saves 95% of the power required by a flat structure. The cost ratio drops significantly as the network scales, from 0.575 at $N=10$ to 0.0447 at $N=2000$.
4. Conclusion and Future Work
This paper provides a rigorous mathematical foundation for hierarchical WSN optimization. While the current model assumes uniform grid deployment and static topology, it establishes clear benchmarks for energy savings. Future research will extend these findings to multi-level hierarchies and irregular 3D deployment environments to enhance real-world applicability.
Summary updated