A Dynamic Wireless Spectrum Allocation Algorithm for Sliding Scheduled Demands

Spectrum management in cognitive radio is one of the most challenging issues and has attracted much research attentions in recent years. However most of the existing works consider either online demands or scheduled demands, which may not fully capture the flexible nature of the demands in certain circumstances, such as the communication demands for a synchronized sensing data transmission in wireless sensor networks which may not require the wireless channels be assigned in a pre-specified time point and could instead be provisioned in a flexible time range. Sliding scheduled traffic model, which allows the starting time of the service provisioning for a demand to slide in a pre-specified time window, instead of a time point, fits in well. Based on this observation, we propose in this paper to model the wireless channel demands for communication with sliding scheduled traffic model to improve the utilization ratio. We presents a mixed integer linear program (MILP) formulation as well as an efficient heuristic algorithm for this optimization problem. Numeric results show that adopting sliding scheduled demand model significantly improves the utilization ratio of wireless channels in dynamic wireless spectrum allocation.