CRITICALITY-AWARE LOCALIZATION FOR ENERGY-EFFICIENT WIRELESS SENSOR NETWORKS

Abstract

This paper addresses the fundamental energy-accuracy trade-off in wireless sensor network localization through a novel criticality-aware framework. Unlike conventional approaches that employ uniform localization strategies across all nodes, our method introduces a dynamic, multi-faceted criticality assessment that continuously evaluates each node based on application-defined priority, residual energy levels, and environmental data gradients. Nodes are autonomously classified into three distinct tiers—critical, semi-critical, and non-critical—enabling tier-specific localization strategies: high-precision Time Difference of Arrival (TDoA) with frequent updates for critical nodes, Received Signal Strength Indicator (RSSI) with Kalman filtering for semi-critical nodes, and energy-efficient centroid localization for non-critical nodes. The framework incorporates a Bayesian-inspired criticality model with parameters optimized through extensive grid search analysis, ensuring optimal balance between localization precision and energy conservation. A distributed coordination protocol maintains global energy constraints through adaptive threshold adjustment while guaranteeing convergence to Nash equilibrium. Comprehensive simulation results demonstrate that our approach maintains exceptional accuracy for critical nodes (0.92 ± 0.05 m average error) while reducing energy consumption for non-critical nodes to 0.15 J per update. Compared to state-of-the-art approaches, our method achieves a 42% extension in network lifetime while maintaining statistical significance across all performance metrics (p < 0.01). The proposed solution provides a principled, scalable approach to resource-aware localization that adapts to dynamic network conditions and application requirements.