Optimized Resource Allocation for Energy Efficiency in Cognitive Radio Networks: A Comparison of Particle Swarm Optimization and Whale Optimization

The proposed research focuses on the optimal utilization of energy resources in a cognitive radio system using the Particle Swarm Optimization technique, comparing the Whale Optimization method for better accuracy. This work makes use of the Kaggle dataset considered as a benchmark for research studies in the sphere of resource allocation. For this purpose, two different groups were considered for Whale Optimization algorithm performance (Group 1) and Particle Swarm Optimization method performance (Group 2), with 20 samples in each group. Thus, the total sample size obtained was 40. This number was determined a priori to ensure the statistical robustness of the results. These statistical analyses were performed using IBM SPSS software, aspiring for 80% statistical power. It does this in order to avoid both Type I and Type II errors: A significance threshold of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\alpha=0.05$</tex> and a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\beta$</tex> The values used were 0.2 for the significance and 95 % confidence level. The accuracy of Whale Optimization was 84.15 % and that of Particle Swarm Optimization was 94.12 %, showing a very impressive improvement in accuracy. Then, using a two-tailed test of significance returned a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{p}$</tex>-value of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{0. 0 5 5}$</tex>, which is less than the preset significance level of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{0. 0 5}$</tex>. From this, it can be deduced that Particle Swarm Optimization indeed performs effectively in improving the accuracy of resource allocation in computer networks.