An improved cell type classification for random walk modeling in cellular networks

Abstract

In an earlier paper by Akyildiz et al, it is shown how to classify cell types in a cellular network based on the random walk model; the number of states is reduced from a naive classification of (3n(2) + 3n - 5) to n(n + 1)/2 in a hexagonal configuration, where n is the number of layers of cells. By using a reflection relation, this paper shows that the number of states can be further reduced to (n + 1)(n + 3)/4 if n is odd, and n(n + 4)/4 if n is even. These numbers are about half of that of Akyildiz et al. Simulation experiments indicate that our approach significantly reduces the computational costs in the related probability derivation.