Research On Key Technologies In Cognitive Radio

With the development of communication technologies, the communication scheme has transformed from narrow-band system with low speed text transimission to wireless broadband system that transfers high speed multimedia information. Thus, the requirement for spectrum resource of wireless communication is increasing dramatically nowadays while the available spectrum resource is however very limited. The present resource allocation method employs fixed distribution according to communication service type, leading to insufficient spectrum resource leftover. Dynamic resource allocation can greatly improve the spectrum utilization, while cognitive radio is the key technology to implement it. So it is of importance to research on cognitive radio technology.Cognitive radio is an intelligent wireless communication system based on software radio. By observing the surroundings, it adaptively adjust internal system parameters (e.g. transmitting power, carrier frequency and modulation mode) according to signals received from RF in order to achieve a better communication performance.This paper focuses on some key technologies in cognitive radio including local spectrum sensing, cooperative spectrum sensing, malicious user detection, sensing parameter optimization, channel estimation and spectrum resource allocation scheme, etc.Spectrum sensing is the one of the key technologies in cognitive radio. Spectrum sensing technology with superior performance can improve the communication capacity of cognitive users and provide sufficient protection on primary users. Basic concepts of local spectrum sensing and energy detection which is very popular in local spectrum sensing are introduced in charpter 2 firstly. To solve the sensing performance degradation of energy detection that exists in noise uncertain environment, local spectrum sensing algorithm with ability to resist noise uncertainty is proposed. The algorithm can agilely change system false alarm probability by inducing an adaptive threshold parameter p. It is a practical algorithm that does not require any prior knowledge of primary user's signals and noise variance compared to energy detection. Because of the hidden node problem and shadowing effect, the local spectrum sensing results are not reliable. Cooperative spectrum sensing is proposed to solve this problem. The basic concept of cooperative spectrum sensing is briefly introduced in charpter 2. Then, an improved D-S detection algorithm is presented that assigning weights to users according to their SNR values. By making use of those weights of users, the impact of users with unreliable detection results is weakened in the fusion center. A malicious user detection algorithm is proposed to effectively detect malicious users who always report 0 or 1 as well as the hidden malicious users that report 0 or 1 with arbitrary probability.The tradeoff relationships in spectrum sensing are studied. Spectrum sensing research status is introduced firstly. There are two purposes to optimize sensing parameters, one is improving detection performance, and the other one is optimizing user's throughput and spectrum efficiency. The relationships between sensing time and data transmitting time, sensing time and system spectrum efficiency are investigated. The optimal local sensing time which can provide sufficient protection on primary user and more data transmitting time for cognitive users is derived. The relationship between cooperative sensing time and system spectrum efficiency is presented. Cooperative spectrum sensing can improve sensing performance. It will spend less time to achieve the target detection probability, while more time is left to transmit data.After finding the spectrum access opportunities, cognitive users need to inspect the activity of primary user, provide sufficient protection on primary users and they should sufficiently exploit the spectrum opportunities to communicate in high efficiency. This requires high efficient communication technologies with low computational complexity. OFDM is a high spectrum efficiency communication scheme and also it is taken into consideration to be used in IEEE 802.22. According to these, an OFDM channel estimation algorithm with low complexity and high performance is proposed in chapter 4. By choosing smoothing parameterλ, it obtains the optimal estimated channel without knowing any prior knowledge of users. The proposed algorithm has similar computational complexity as liner interpolation while its bit error rate performance is much better than liner interpolation and is close to MMSE.Finally, the problem of spectrum resource allocation in OFDM based cognitive cellular network is investigated in chaper 5. First of all, the primary spectrum resource allocation method in traditional cellular network as well as its pros and cons are discussed. Then the next generation of wireless communication and dynamic spectrum multiplexing in LTE are introduced. Lastly, the spectrum allocation solution for uplink and downlink in cognitive cellular network is proposed. By making use of free TV band, the spectrum efficiency and throughput of edge users in network are enhanced so that the throughput of entire network is improved.