Simulation studies on scheduling requests for multiple data items in on-demand broadcast environments

Data scheduling for on-demand broadcast has been studied extensively in the past few years. Most work has focused on single-item requests in single-channel architectures. However, this model is not practical for emerging mobile applications. This study investigates data scheduling for multi-item requests in multi-channel on-demand broadcast environments. The performance of several well known scheduling algorithms designed for single-item requests in this new environment is comprehensively evaluated. Two problems, namely, request starvation and broadcast mismatch, are identified. The results showed that existing algorithms cannot perform efficiently in this new setting and it is imperative to develop new solutions to overcome the problems.