In this thesis, spectrum sensing techniques are investigated for cognitive radio (CR) networks in order to improve the sensing and transmission performance of secondary networks. Specifically, the detailed exploration comprises of three areas, including single-node spectrum sensing based on eigenvalue-based detection, cooperative spectrum sensing under random secondary networks and full-duplex (FD) spectrum sensing and sharing techniques. In the first technical chapter of this thesis, eigenvalue-based spectrum sensing techniques, including maximum eigenvalue detection (MED), maximum minimum eigenvalue (MME) detection, energy with minimum eigenvalue (EME) detection and the generalized likelihood ratio test (GLRT) eigenvalue detector, are investigated in terms of total error rates and achievable throughput. Firstly, in order to consider the benefits of primary users (PUs) and secondary users (SUs) simultaneously, the optimal decision thresholds are investigated to minimize the total error rate, i.e. the summation of missed detection and false alarm rate. Secondly, the sensing-throughput trade-off is studied based on the GLRT detector and the optimal sensing time is obtained for maximizing the achievable throughput of secondary communications when the target probability of detection is achieved. In the second technical chapter, the centralized GLRT-based cooperative sensing technique is evaluated by utilizing a homogeneous Poisson point process (PPP). Firstly, since collaborating all the available SUs does not always achieve the best sensing performance under a random secondary network, the optimal number of cooperating SUs is investigated to minimize the total error rate of the final decision. Secondly, the achievable ergodic capacity and throughput of SUs are studied and the technique of determining an appropriate number of cooperating SUs is proposed to optimize the secondary transmission performance based on a target total error rate requirement. In the last technical chapter, FD spectrum sensing (FDSS) and sensing-based spectrum sharing (FD-SBSS) are investigated. There exists a threshold pair, not a single threshold, due to the self-interference caused by the simultaneous sensing and transmission. Firstly, by utilizing the derived expressions of false alarm and detection rates, the optimal decision threshold pair is obtained to minimize total error rate for the FDSS scheme. Secondly, in order to further improve the secondary transmission performance, the FD-SBSS scheme is proposed and the collision and spectrum waste probabilities are studied. Furthermore, different antenna partitioning methods are proposed to maximize the achievable throughput of SUs under both FDSS and FD-SBSS schemes.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:738947 |
| Date | January 2017 |
| Creators | He, Yibo |
| Contributors | Thompson, John |
| Publisher | University of Edinburgh |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/1842/28951 |
Page generated in 0.0023 seconds