On the Controller Placement Problem: a New SD-IoT Topology for the Algerian Network

International audience In disaster management, Software Defined Internet of Things (SD-IoT) plays a critical role by enabling real-time data collection and network adaptability, which are essential for effective response and recovery. One of the key challenges in SD-IoT is controller placement, which significantly impacts network latency and overall performance. In this paper, we deal with the controller placement issue and propose a new SD-IoT-based infrastructure topology for the Algerian Network. Using the K-means clustering algorithm, we partition the network topology into different clusters (domains), each managed by a one controller. Then, we propose an enhanced version of K-center and K-median algorithms tailored for a weighted topology whose flows are proportional to the state’s population. This ensures that controller placement is optimized based on the population distribution, enhancing the overall efficiency and responsiveness. Simulation results show that our population-aware SD-IoT infrastructure provides optimal placements for controllers that reduce latency, and so improve the resilience and effectiveness of disaster management.