DSNs Coverage Optimization Based on Improved Multiobjective Army Ant Search Optimizer

Directional sensor networks (DSNs) are a vital branch of wireless sensor networks. In DSNs, area coverage optimization is still a critical issue, and when considering the network lifetime, researchers usually treat the network lifetime as a constraint, and achieve area coverage optimization by moving or rotating sensors that have a sensing angle between 0 and 2
$\pi $
. However, optimizing both coverage and network lifetime at the same time has not yet been thoroughly investigated. Since the number of active nodes can directly affect the network lifetime, we propose the improved multiobjective army ant search optimizer (IMOAAO) to optimize the sensing direction of the sensors and their active states, ultimately achieving higher network coverage rate with fewer active sensors. The IMOAAO is based on the army ant search optimizer (AAO) with the following four added improvements. Firstly, fast-nondominated sorting is proposed in NSGA-II, mainly used to transform single-objective algorithms into multiobjective algorithms. Secondly, chaotic mapping is utilized to initialize the population to increase population diversity. Afterward, the average fitness value adaptor is introduced to determine whether the algorithm falls into local optimal. Finally, a multiobjective competitive swarm optimizer (MOCSO) is added to increase the optimization ability. The simulation results show that IMOAAO has a better performance under mutiobjective test functions compared to some best known existing multiobjective optimization algorithms. When applying to DSNs coverage optimization, IMOAAO can also explore a better Pareto frontier solution, after analyzing the complexity, it was concluded that IMOAAO at the expense of some of the complexity greatly improves the ability to achieve coverage optimization, and can provide a wider range of optimal choices for the decision maker to choose from.