A neural network and genetic algorithm scheme for optimal dynamic channel assignment in mobile networks

Channel assignment problems (CAP) are classified as NP-hard. The major challenge is how to assign the limited channels optimally while ensuring that electromagnetic interference is minimized and an acceptable quality of service (QoS) is provided. Inefficient channel assignment leads to congestion in mobile networks. This work proposes a scheme that combines the easy convergence ability of neural network and the global search ability of genetic algorithm to address the CAP. The proposed solution to the CAP was tested on Philadelphia standard benchmark problems of frequency assignment and produced a lower call blocking probability at high traffic load thereby proving to be more efficient compared to other schemes.