The GA Solution of Dynamic Spectrum Allocation in Cognitive Radio Based on Collaboration and Fairness

This paper proposed a GA (Genetic Algorithm) based algorithm to solve dynamic spectrum allocation through analyzing the current theory and using the mathematical model-matrices of available spectrum, utilization (bandwidth or throughput) and interference. According to the principle of collaboration and fairness, we take the multi-objective of maximizing the system utilization and fair assignment as initial heuristic rule and fitness function in the global optimal searching. Experimental results confirm that GA solves the local-optimum problem of conventional algorithm. The optimization rate can reach 10%-20%. Meanwhile the space for searching is effectively compressed applying the heuristic rule and convergence is acceptable.