Optimal Placement of UAVs Forming Aerial Mesh Networks to Handle Network Issues
Manali Gupta and Shirshu Varma
The optimal placement of UAVs to facilitate post-disaster emergency communication services is a paramount research domain. The novelty of our work is to optimally place available UAVs in 3D space to meet the objectives prominent during such situations. The objectives considered here are target coverage, QoS, total energy consumption, and newly characterized objectives, i.e. network robustness and fault resistant, to efficiently handle the events of node failure/draining out. These objectives conflict with each other while positioning the UAVs. To efficiently tradeoff these conflicting objectives, we proposed two metaheuristic based hybrid optimization algorithms, namely, HEPO-HS, a hybrid of Emperor Penguin Optimizer (EPO) and Harmony Search (HS); and HEPO-PSO, a hybrid of EPO and Particle Swarm Optimization (PSO). HEPO-HS improves the exploitation ability of EPO using harmonics based crossover of HS, while HEPO-PSO improves the exploitation ability of EPO using swarm-based position update of PSO. Further, to improve network performance, we tuned the parameters of the hybrid algorithms using well-known Taguchi’s design of the experiment. Also, maintaining network connectivity during such emergencies is necessary, therefore, we used the concept of connected components for graphs. The proposed hybrid algorithms are compared with the existing popular optimization algorithms like genetic algorithm (GA), PSO, HS and EPO over different test scenarios, i.e. small-scale, medium-scale, and large-scale. ANOVA test and Tukey’s posthoc test suggest that in small-scale and medium-scale scenarios, the proposed HEPO-HS significantly performs better than most of the existing popular algorithms i.e. GA, PSO, HS and EPO. However, in large-scale, the proposed HEPO-PSO significantly performs better than the existing popular algorithms. Moreover, being a hybrid, the computation time of hybrid algorithms is more than the existing algorithms.
Keywords: Genetic Algorithm, metaheuristic optimization, Taguchi’s design of experiment, connected components, minimum vertex cut