Efficient Source Detection in Incomplete Networks via Sensor Deployment and Source Approaching

Le Cheng, Peican Zhu, Keke Tang, Chao Gao, Zhen Wang

Abstract: Rumor source detection in structurally incomplete networks holds significant practical importance. Existing methods predominantly assume a complete network structure information; furthermore, they often neglect the issue of resource consumption, i.e., sensor deployment. In this paper, we propose an efficient source detection approach in incomplete networks via propagation-aware Sensor Deployment and time stamp-guided Source Approaching (SDSA) to tackle these challenges. Specifically, during the sensor deployment phase, we employ quality-guaranteed Monte Carlo propagation simulations coupled with a greedy strategy to achieve maximum coverage with minimal sensors. In the source detection phase, for the structurally incomplete network snapshots, we first attempt edge reconnection from the sensor with the earliest timestamp, followed by posterior maximization Bayesian estimation for source identification. Extensive experiments demonstrate the effectiveness of SDSA and its superiority over state-of-the-art methods. The code has been made publicly available at https://github.com/cheng-le/SDSA.