Advanced squirrel algorithm-trained neural network for efficient spectrum sensing in cognitive radio-based air traffic control application

Abstract

In the current scenario, there is a drastic increase in air traffic. The air to ground communication plays a crucial role in the air traffic control system. There is a limited spectrum available for aircraft to establish a connection with the Air Traffic Controller (ATC). With air traffic growth, the available spectrum is getting more congested. This paper proposed an Advanced Squirrel Algorithm (ASA)-trained neural network (NN) for efficient spectrum sensing for cognitive radio-based air traffic control applications. ASA is a novel metaheuristic-based training algorithm for an NN. With the proposed algorithm, it is possible to dynamically allocate the unused spectrum for air to ground communication between aircraft and ATC. The quantitative analysis of the proposed ASA-NN-based spectrum sensing is done by comparing it with the existing metaheuristic-based NN training algorithms, namely, particle swarm optimization Gravitational Search Algorithm (PSOGSA), particle swarm optimization (PSO), gravitational search algorithm (GSA), and artificial bee colony (ABC). Simulation-based evaluation shows that the proposed ASA-NN is capable of efficiently detecting the spectrum holes with high convergence rate as compared to PSOGSA-, PSO-, GSA-, and ABC-based algorithms.