Multimedia Channel Assignment in Cellular Networks

This paper deals with the channel assignment problem in a hexagonal cellular network with two-band buffering, supporting multimedia services. We consider the simplest case of a multimedia cellular network dealing with only two different types of multimedia signals, where each cell has a single demand for each type. We first derive the lower bounds on the minimum bandwidth requirement for assigning multimedia channels to a seven-node subgraph of the hexagonal cellular network. We next estimate the lower bounds on the minimum bandwidth requirement for assigning channels in some real-life situations where the relative values of frequency separation constraints are somewhat restricted. Next, we present an algorithm for solving the multimedia channel assignment problem in its most general form, using Genetic Algorithm (GA). We then propose a technique for a clever re-use of the channels, by exploiting the hexagonal symmetry of the cellular network and using only eighteen distinct frequency bands on a nine-node subgraph of the network. With this concept of re-using the channels, we first find the required frequency separation constraints among the channels to be assigned to the different nodes of the network, and then use our proposed GA-based algorithm for assigning the multimedia channels for the complete network. Experiments with different values of the frequency separation constraints show that the proposed assignment algorithm converges very rapidly and generates near-optimal results with a bandwidth pretty close to our derived lower bound.