Spectrum Allocation Technology Based On Game Theory In Cognitive Radio Networks

As the development of telecommunication technology, the lack of radio spectrum has become a bottleneck of wireless communication and related services. However, this problem, indicated by former researches, is due to traditional rigid spectrum allocation scheme which leads to most of the scarce spectrum being idle for long time. Cognitive radio technology which could sense the usage of around radio spectrum is a promising technology to solve the problem and improve the efficiency of spectrum usage. As a result, dynamic spectrum allocation in cognitive radio networks has become a research hot point all round the world. Game theory became one of the most important tools in the research because algorithms based on it could fully exploit the sensing ability of wireless devices which will result in high spectrum efficiency.First of all, the paper deeply discusses game model, analyzes the Nash equilibrium and its solution, and introduces the optimal mechanism design theory. Besides, this paper generalizes current spectrum allocation model for cognitive radio networks include models based on auction theory, game theory models, graph coloring models and interference temperature models.In addition, focusing on multi-user OFDMA system, this paper proposes a radio resource allocation algorithms based on Nash Bargain Solution about two users, then allies Hungry tasks assignment algorithms to generalize this algorithms to multi-user scenarios. The algorithm makes a tradeoff between overall system throughput and fairness. Tested through Matlab, the overall system performance is better than maximized overall rate algorithms and maximized minimal rate algorithms. Its complexity is O(K⁴ + K² N log₂N), which is lower than other algorithms with similar performance.Finally, this paper creatively introduces the waiting-line auction theory to algorithms design for wireless networks. It proposes a new spectrum access scheme and algorithm. This algorithm introduces primary user utility which effectively reduces the algorithm complexity. It not only maximizes the utility of primary user but also optimizes the network throughput. Besides, this paper applies the auction theory to WiFi networks. It proposed network access algorithm in WiFi networks. This algorithm adjusts the users' content windows based on users' service load and the usage of spectrum around. Compared with original WiFi networks via OPNet, the proposed algorithm increases the overall thought.