One of the most important designing problems in cellular communication is the Channel ASSIGNMENT Problem (CAP). The goal of channel ASSIGNMENT is to assign the required number of channels to each cell, such that FREQUENCY spectrum utilization is maximized and FREQUENCY interference effect is minimized. Due to the limitation of the usable FREQUENCY spectrum, the optimal channel ASSIGNMENT can improve capacity, quality, and performance of the system. Since in its simplest form this problem is known to be equivalent to a graph-coloring problem, belonging to the class of NP-complete optimization problems, different techniques such as heuristic techniques, neural networks, simulated annealing, and evolutionary algorithms have been used to solve it. In this paper, a modified Hopfield neural network has been proposed. The modifications, include structure of the network, initialization method, and the updating procedure. In the network structure, contrary to usual structures, there is also a connection from each neuron’s output to its input with negative weight and instead convergence is obtainable with the procedure of updating. Simulation results show better performance than the previous presented networks.