A data-driven robustness algorithm for the Internet of things in smart cities

The Internet of Things has been applied in many fields, especially in smart cities. The failure of nodes brings a significant challenge to the robustness of topologies. The IoT of smart cities is increasingly producing a vast amount different types of data, which includes the node's geographic information, neighbor list, sensing data, and so on. Thus, how to improve the robustness of topology against malicious attacks based on big data of smart cities becomes a critical issue. To tackle this problem, this article proposes an approach to improve the robustness of network topology based on a multi-population genetic algorithm (MPGA). First, the geographic information and neighbor list of nodes are extracted from a big data server. Then a novel MPGA with a crossover operator and a mutation operator is proposed to optimize the robustness of topology. Our algorithm keeps the initial degree of each node unchanged such that the optimized topology will not increase the energy cost of adding edges. The extensive experiment results show that our algorithm can significantly improve the robustness of topologies against malicious attacks.