Evolutionary neural networks algorithm for the dynamic frequency assignment problem

TL;DR: A new approach for solving the problem of frequency allocation based on using initially a partial solution respecting all constraints according to a greedy algorithm, which shows more efficiency in terms of flexibility and autonomy.

Abstract: Wireless communication is used in many different situations such as mobile telephony, radio and TV broadcasting, satellite communication, wireless LANs, and military operations. In each of these situations a frequency assignment problem arises with application-specific characteristics. Researchers have developed different modelling ideas for each of the features of the problem, such as the handling of interference among radio signals, the availability of frequencies, and the optimization criterion. This paper presents a new approach for solving the problem of frequency allocation based on using initially a partial solution respecting all constraints according to a greedy algorithm. This partial solution is then used for the construction of our stimulation in the form of a neural network. In a second step, the approach will use searching techniques used in conjunction with iterative algorithms for the optimization of the parameters and topology of the network. The iterative algorithms used are named hierarchical genetic algorithms (HGA). Our approach has been tested on standard benchmark problems called Philadelphia problems of frequency assignment. The results obtained are equivalent to those of current methods. Moreover, our approach shows more efficiency in terms of flexibility and autonomy.