Graph Theoretic Approach to QoS-Guaranteed Spectrum Allocation in Cognitive Radio Networks

We propose a novel graph theoretic matching algorithm to model the dynamic channel allocation problem in cognitive radio networks. In the past, no research work has considered both the channels and the user request priorities in the context of spectrum allocation. In our work, we define a unique model to assign priorities among the channels in a spectrum based on the transmission power allowed in each channel to mitigate interference on the licensed primary users. Crisp logic lines categorize requests into distinct priorities based on the user requirements. These prioritized channels and user requests then constitute the set of vertices V of a bi-partite graph G = (V,E), whose set of edges E represent the communication links between the channels and the users. We propose a novel O(|E|) priority-satisfying matching algorithm to design an allocation scheme to satisfy the quality of service guarantees of the unlicensed secondary users. We evaluate the performance of our proposed algorithm based on critical parameters such as latency and packet dropping probability.