Health-aware sensor network for smart street lighting multi-agent trust mechanisms

In this paper, a smart street lighting architecture that enables municipalities to manage and control public street lamps in the presence of potential sensor faults is considered. By exploiting traffic information, the system can autonomously adjust street lamp brightness in targeted urban areas or road segments, minimizing energy consumption while complying with regulatory limits. To this end, a traffic model, formulated as a compartmental system, is proposed to estimate the number of vehicles in each segment by leveraging sensor networks and state observer techniques. The goal is to solve the problem of estimating the state of a multi-agent system. In particular, at each time instant, the available information is exploited using a reputation mechanism based on multi-agent consensus strategies and Perturb & Observe algorithms. A key feature of this methodology lies in its ability to identify possible faulty sensors by evaluating their Quality of Service (QoS) using trust metrics shared among agents. Finally, the effectiveness of the proposed approach is tested on a highly structured road network to assess its capability in estimating traffic flow and adapting street lamp brightness. In this respect, the simulations show energy savings of up to 70% compared to traditional street lamp management systems.