Multi-Objective Task Allocation for Dynamic IoT Networks

This paper proposes a Multi-Objective Optimization Algorithm for task allocation within dynamic Internet of Things (IoT) networks, which experience both device mobility and device failures. A prediction of future node positions is combined with a diversity-enhancing archiving mechanism, combining and improving upon task allocation optimization algorithms for mobile (Mobility-Aware Multi-Objective Task Allocation Algorithm) and error-prone (Availability-Aware Multi-Objective Task Allocation Algorithm) networks. Additionally, an elitism mechanism serves to preserve the most optimal solutions. The proposed algorithm is thus capable of optimizing task allocations in IoT networks throughout node failures while simultaneously accounting for future node positions. The proposed algorithm is evaluated on networks with different error rates and amounts of mobile nodes. The evaluation shows the improved performance regarding the metrics network lifetime, latency, and availability.