Modeling And Reserach On Network Selection Strategy In Cognitive Radios

The demand for wireless network service has increased dramatically in recentdecades, which calls for more radio frequency resources. Spectrum resources have beenstatically regulated by governmental agencies. Many studies have shown that more than90%of the exclusively allocated spectrum is underutilized. Cognitive radio is a smarttechnology to solve this problem, which exploits vacancies in licensed spectrum intemporal and geographic domains and admit secondary users use these bands forunlicensed transmissions without causing interference to the primary users.Dynamic spectrum sharing is a promising approach to reusing the underutilizedradio spectrum in a cognitive radio system. Choosing which primary network to accessfor secondary user brings the network selection problem. Potential available spectrumfor CRS can be from multiple primary radio systems with different spectrum prices, andsecondary users with heterogeneous usage modes, interference punishments and servicetypes. This scenario generates new opportunities for cognitive radio system to improvethe system performance by exploiting heterogeneous characteristics of primarynetworks.In this paper, a novel based on service types of SU network selection strategy isproposed. Continuous time Markov chain model is used to describe the networkselection problem and the performance metrics are derived. Then Random and Greedynetwork selection strategies are proposed under this framework. Motivated by thedrawbacks of those strategies, we categorize the service types of SU into Real-TimeServices and Non-Real-Time Services and propose an optimal network selectionalgorithm to maximize performance metrics based on policy iterations.Lastly, the performance of the proposed designs is compared and evaluated throughsimulation and numerical results. The result demonstrates that the proposed novelnetwork selection strategy can satisfy the QoS requirement of Real-Time Services andNon-Real-Time Services and outperforms Random strategy and Greedy strategy.