Research On Spectrum Access And Allocation In Cognitive Radio System

With the advent of the mobile Internet era, the massive new applications begin to emerge and the scarcity of radio spectrum is highlighted. The unallocated radio frequency resources become fewer. However, studies have shown that, currently radio frequency resources based on the authorized allocation are poor in utilization. Cognitive radio improves the existing spectrum utilization through dynamic spectrum access and provides high-speed, reliable communications services for cognitive users. But cognitive user and the authorized user status is unequal. Cognitive user can not interfere with the authorized user of communications, while as much as possible to avoid the interference from the primary user. This requires that the access control for each user in the cognitive system and power control to control interference for high-speed and reliable data transfer. Thus access control technology research is significant for cognitive radio system. On the other hand, for optimal allocation of the primary users of shared spectrum, how to efficiently use spectrum resources, which is also the core issues in the cognitive radio system. This paper mainly studies the problem of access control and spectrum sharing in cognitive radio system, so as to improve the theory of dynamic spectrum access for cognitive radio systems. The main work and contributions are as follows:(1) A user admission and power control algorithm for OFDM-based cognitive radio is proposed.Secondary users (SUs) may interfere on primary users (PUs) for OFDM-based cognitive radio systems due to the non-orthogonality of their respective transmit signals. First, the number of admitted secondary links is maximized, while the interference introduced to every PU remains within its own tolerable range and the SNR requirement of all SUs are met. It is proved to be an NP-hard problem. A suboptimal algorithm is proposed that removes the SU who contributed most interference to some PU, according to the distribution of interference. Second, power control for the admitted SUs are investigated in order to maximize the total transmission rate of the SUs, then an iteration algorithm with low complexity is proposed by the lagrangian primal-dual approach. Simulation results show that the effectiveness of the optimizing.(2) Access parameters optimization problems for cognitive radio systemFirst, an access parameter optimization solution for interference model based on Bayesian risk is studied. According to the tolerance of system about the probability of false alarm, the probability of false detection, their weighting is defined as Bayesian risk. Then the function between the transmission time and the interference for given Bayesian risk value is derived. On this basis, for given the Bayesian risks, when meeting interfere constraints, under the non-cooperative sensing mode, the joint optimization of decision threshold, sensing time and data transmission time are investigated; under cooperative sensing mode, a design algorithm about sensing time and data transmission time is proposed, which is based on Newman-Pearson criterion. Simulations results show, even if the Bayesian risk value is large, as optimization of the frame length, the proposed algorithm can also achieved better throughput compared with the existing schemes. Second, in order to improve the secondary user’s access time, a frame length design algorithm which limits interference and spectrum opportunity loss is proposed. When a false alarm occurs or properly detects that the primary user occupies the channel, but the primary users leave in the transmission time, on the same time, cognitive users don’t access the channel, which will lead to lost spectrum opportunity for congitive users; when mistaken detection occurs or correctly detects that the primary user’s channel is idle, but the primary user access the channel in the transmission time, which will produce interference. Under the exponentially distributed assumption based on the primary user activity, an interference model and a lost access opportunity model for secondary users are derived. The models ensure interference time and lost access time is less than a specified percentage with a high probability. Then a theoretical framework about frame duration is developed to limit interference and lost access opportunity ratio. Further simulation shows the frame structure developed in such a way can effectively ensure the secondary user’s access time, the higher sensing efficiency and lower interference rates.(3) A calculation method of data transmission time for cognitive radio system is proposed.Under exponentially distributed assumption based on the primary user’ channel occupancy/idle time, an interference model about transmission time, in which the frequency of occurrence of primary user is unlimited, is derived and then a scheme that is used to determine transmission time is developed. Under this transmission time, the primary user interference can be accurately controlled. Further simulation shows the experimental value of the interference rate is very close to theoretical value of the interference model, but the gap between experimental value and theoretical value is obvious in the existing methods, which verify fully the correctness of the interference model.(4) The pricing algorithms for free spectrum in spectrum trade are proposed.For spectrum sharing scheme based on the spectrum markets, pricing is a difficult problem, aiming at pricing for idle spectrums in a cognitive radio network with multiple primary users, In the competitive pricing, based on non-cooperative game theory, the authors propose a competitive pricing solution algorithm and prove the algorithm converges to a unique Nash equilibrium; In the cooperative pricing, by solving the dual problem of the original problem, propose a cooperative pricing solution algorithm and prove that the algorithm converge to the global optimal solution if the step size is small enough. The simulation results show that the two proposed algorithms run faster and get more total revenue but the revenue values are very close to the optimal solution.