Resilient Optimization of Sensor Networks Deployment Based on Graph Similarity Learning and Node Resilience Prediction

Network resilience, defined as the ability of a network to maintain its functionality under failure or adversarial conditions, is a crucial consideration in the design and deployment of network systems. However, traditional optimization approaches for sensor network deployment often neglect this essential factor. This paper proposes a resilient optimization method for sensor network deployment that integrates graph similarity learning with node resilience prediction, with a focus on prioritizing network resilience. By strategically selecting and incorporating optimal backup nodes, this method maximizes the enhancement of network resilience performance. Experimental results demonstrate that the proposed approach achieves a more efficient distribution of backup nodes, leading to significant improvements in both network robustness and energy efficiency.