Minimal cost coverage planning for single frequency networks

OFDM based single frequency networks (SFNs) have been standardized for terrestrial broadcasting systems, for digital audio broadcasting (DAB) as well as for digital video broadcasting (DVB). Due to the multipath tolerance of the OFDM scheme, the receiver is able to combine signals coming from several transmitters, despite of the varying propagation delays, i.e., heavy artificial multipath propagation. In order to take full advantage of the diversity gain provided by the SFN architecture, proper network design is required. We focus on the cost efficient design of an SFN providing broadcasting services over a predefined service area with requirements both on the received signal quality and on the allowable interference level experienced by existing services in the same spectrum. We formulate the problem as a discrete optimization problem, where the network design parameters such as power, antenna heights and transmitter locations are the decision variables. The general stochastic optimisation algorithm simulated annealing has been adapted for solving the above problem. The novelty of our method is that cost factors and interference constraints are embedded in the optimisation procedure. Through numerical examples we demonstrate that significant reduction in network cost can be achieved by our approach.