Approches Pour La Classification Du Trafic Et L'optimisation Des Ressources Radio Dans Les Réseaux Cellulaires : Application À L'afrique Du Sud Dedication Acknowledgements Declaration
The demand for more efficient and fast channel allocation techniques in cellular systems increases day by day. Borrowing channel assignment (BCA) was introduced in the literature as a compromise between the classic fixed and dynamic channel allocation schemes. This paper examines the behavior of three heuristic BCA techniques based alternatively on a Hopfield neural network, an efficient evolutionary algorithm named combinatorial evolution strategy (CES) and a third heuristic which combines the basic advantages of the two above computational intelligence methods. By considering some specific assumptions that follows an ideal cellular mobile model, BCA is formulated as a combinatorial optimization problem. The above heuristics have been extensively applied to solve efficiently such problems in the past. Simulation results, derived for uniform and nonuniform traffic load conditions, are used to compare these BCA schemes each other as also with other well-established allocation techniques.