An Improved Genetic Algorithm for Bi-Level Multi-Objective Q-Coverage in Directional Sensor Networks

Direction sensor networks are robust systems employed for detecting phenomena in environments or monitoring objects therein. They have a wide range of applications across many different industries and fields. In terms of the availability of resources, direction sensor networks deal with two problems: over-provision and under-provision of sensors. Over-provision occurs when there are too many sensors in the monitoring area, resulting in wasted resources and unnecessary energy consumption as some sensors are not well utilized. In contrast, under-provision occurs when there are too few sensors in the monitoring area, leading to the coverage of targets not satisfied. To ensure balanced coverage in under-provisioned environments, sensors must be placed so as to provide nearly equal fault tolerance to all objects, thereby enhancing the operational efficiency of the network. On the other hand, in over-provisioned environments, the number of active sensors needs to be minimized so that energy consumption is efficient. This study focuses on solving the Q-coverage problem in adjustable-orientation direction sensor networks, aiming to optimize a bi-level objective: maximizing network coverage balancing while minimizing sensor count in both under-provisioned and over-provisioned environments. The proposed Improved Genetic Algorithm utilizes novel operators, including Greedily-tuned Simulated Binary Crossover and Adaptive Polynomial Mutation. Evaluation parameters, including the Q-Balancing Index, Distance Index, Coverage Quality, Power Consumption, and the number of active sensors, demonstrate the efficiency of the proposed algorithm compared to other existing methods.