Network Selection Of Cognitive Radio Network With Multiple Licensed Networks

Cognitive radio networks(CRNs)is envisioned to be an effective approach to increase spectrum utilization.In CRNs,the cognitive u sers detect idle spectrum by performing spectrum sensing and then opportunistically tap into underutilized licensed spectrum units,as long as they affect communication activities of PUs insignificantly.However,due to preemptive priority of primary users and primary networks' dynamics,Qo S of cognitive users is bad,which cannot give people good service.Improving Qo S of cognitive users is the key point.Now,thanks to the development of network fusion technology,it is envisaged that multiple primary networks will be incorporated into CRNs,which can offer CRNs varied choices to access different spectrum resources.However,when cognitive users request for access,how to choose the primary network is the question of network selection.In this paper,based on the previous research results,we carry out the research of CRNs with multiple primary networks.We suppose the cognitive users will share the spectrum for a while with primary users rather than drop when cognitive users collide with primary users.To simulate the user behavior,we model the system as a continuous-time Markov chain(CTMC),and derive system performance metrics,for example blocking rate,dropping rate,effective throughput.We also define the interference quantity that is the accumulated interference time to the ongoing primary user.Then based on the above performance metrics,we present a utility function for the CRNs from the point of benefit and punishment.At last,we parameterize network selection strategy in the formulation.Then we discuss three network selection schemes,including Random,Greedy and Weighted strateg ies.The result shows that the weighted strategy performs better and can load balancing.In the primary network,there are many dynamic spectrum accesses(DSA).To improve the flexibility for CRNs,we propose the CRN with multiple primary networks,where primary users have a few DSA strategies.We parameterize network selection strategy in formulation and discuss three network selection schemes,including Random,Greedy and Weighted strategies.The result shows that the weighted strategy performs better and can load balancing.Compared to the CRNs with multiple primary network t hat just have one DSA strategy,the network that have more DSA strategies have better performance.