Decomposition based multiobjective spectrum allocation algorithm for cognitive vehicular networks

To solve the problem of low throughput and fairness in cognitive vehicular networks, we establish a multi-objective spectrum allocation model to maximize the average throughput and fairness of cognitive vehicular networks. By blending the cuckoo search algorithm (CS) with multi-objective optimization based on decomposition (MOEA/D), the multi-objective problem is decomposed into a series of scalar sub-problems with different weights. We verify the effectiveness of the algorithm by comparing it with the MOEA/D algorithm which generates new solutions by cross and mutation on standard test functions ZDT1, ZDT2 and ZDT3, the convergence curve demonstrates that the proposed algorithm has faster convergence rate and more stable convergence performance. We also simulate the spectrum allocation problem model of maximizing the average throughput and fairness of the cognitive vehicular network. The simulation results show that the solution obtained by the algorithm can dominate the optimal solution of some classical algorithms, that is to say this algorithm can improve the network average throughput and fairness at the same time.