Research On Spectrum Sensing For Cognitive Radio Systems

With the rapidly increasing requirement of wireless service, the scarcity of spectrum resource is emerging in recent years. The spec-trum shortage is mainly due to the traditional fixed spectrum allo-cation policy but not the lack of spectrum itself. In order to resolve the contradiction between the spectrum scarcity and inefficient usage, the new spectrum utilization technology should be studied. As an intelligent method to share the spectrum, cognitive radio can sense the spectrum information in temporal, frequency and spacial domain, automatically search and utilize these idle licensed frequency bands, and enhance the spectrum utilization efficiency. The spectrum sens-ing is a key enabling functionality in cognitive radio. This dissertation mainly focuses on the spectrum sensing problem. The main results are summarized as follows:(1) The local spectrum sensing algorithm is considered. An en-hanced energy detection algorithm for random signals buried in Gaus-sian noise is proposed. By operating an arbitrary positive power of the signal amplitude instead of the second power, this algorithm can be viewed as the generalization of the conventional energy detection algorithm. The probabilities of false alarm and detection for enhanced energy detector are derived in closed-form expressions. The optimal value of power corresponding to the best performance is also cal- culated. Simulation results demonstrate that our method achieves achieves better detection performance than the conventional one, es-pecially in low signal to noise ratio environment.(2) Investigated is the cooperative spectrum sensing strategy in centralized network architecture. Based on the interference analy-sis model and periodic sensing mechanism, the key parameters, such as interference duration, effective communication duration, are firstly analyzed in detail. Then the optimization problem of cooperative spectrum sensing is presented and the optimal sensing parameters are obtained to make a tradeoff between interference avoidance and sens-ing efficiency. Finally the cooperative spectrum sensing strategy is extended to multiple licensed channels scenario, which gives an alter-native solution for the deployment of cooperative spectrum sensing in multi-channel multi-user cognitive radio system.(3) Studied is the cooperative spectrum sensing scheme in dis-tributed network architecture. A novel triggered asynchronous spec-trum sensing scheme is proposed. Based on the cooperative spectrum sensing scenario and the licensed channel usage model, the dissimilar sensing result triggering scheme is designed and the likelihood ratio test is performed to combine the local sensing results. This scheme loosens the requirement for local sensing algorithm, observation dura-tion and sensing period. Simulation results show that compared with other synchronous cooperative spectrum sensing methods, our scheme not only decreases the calculation complexity in data fusion, but also guarantees the performance of cooperative spectrum sensing.(4) Introduced is a novel adaptive cooperative spectrum mech-anism, which adaptively regulates the sensing periods according to li-censed channel-usage characteristic and the number of cognitive users in cooperative spectrum sensing. The spectrum characteristic library is constructed to store the channel-usage characteristic information. Then based on the presented adaptive sensing period mechanism the adaptive spectrum sensing control module develops the sensing sched-ule for each licensed channel. Although this method has slightly poorer performance in terms of loss time ratio and interference time ratio than that of periodic sensing mechanism, it reduces the sensing nodes’ power consumption and signaling overhead caused by the delivery of sensing results, and improves system-level sensing efficiency. Therefore this method is suitable for energy-constrained cognitive user scenarios.(5) Analyzed quantitatively is the coexistence of Radar system and TD-LTE system in 2300～2400MHz spectrum band. Based on the principle of link budget, the minimum required value of path loss is calculated in order to achieve the coexistence of Radar system and TD-LTE system. It provides an important reference value for TD-LTE with the application of cognitive radio technology in the 2300 2400MHz spectrum bands.