Because of the limited frequency spectrum and an increasing demand for cellular radio communication services, the problem of channel assignment becomes increasingly important. The channel assignment problem is to assign frequencies to requested calls in a cellular radio network subject to electromagnetic compatibility constraints such that required bandwidth is minimized. In this paper, we propose a new approach to solve the channel assignment problem by modifying the dynamic programming algorithm. We use node color ordering of cells and row wise ordering of calls. The performance of the proposed method is evaluated by solving six well-known benchmarks in channel assignment problems. Results show that this method can find more solutions with minimum required bandwidth in comparison with the other algorithms investigated in this paper. Advantages of our method are number of obtained solutions, independency of generating an initial random solution, and convergence