Abstract:

Cognitive radio has been recently introduced to enhance the spectrum efficiency for it can use the hole of spectrum through sensing the surrounding wireless environment.Most of existing spectrum allocation algorithms for cognitive radio network （CRN）are based on continuous spectrum assignment and the user demands are not considered.For this reason，many small spectrum fragments that are smaller than the user demands are generated and these spectrum fragments can not be fully utilized which make a wasteful use of spectrum resources.In this paper，we introduce an improved spectrum allocation algorithm taking user requirements and discontinuous spectrum aggregation into consideration based on the graph-coloring theory.In the proposed algorithm，the discontinuous small spectrum fragments could be aggregated to meet the user demands as much as possible to enhance the spectrum efficiency.

Key words: cognitive radio, spectrum allocation, user demand, discontinuous spectrum