The Research On User Scheduling And Physical-layer Security Scheme In MIMO Cognitive Radio Systems

With the sharp increase of the number of wireless terminals, the intense competition for spectrum resource will occur soon. Cognitive radio is a promising technology, which allows the coexistence of PU and SU to improve the spectrum utilization. MIMO is also regarded as an advanced technology for its considerable capacity and highly receiving performance. MIMO will be applied in cognitive radio network in the future. The thesis focuses the research on MIMO technology in cognitive radio networks, including two parts:1.the scheme of multi-MIMO SU scheduling; 2.Scheme for preventing legtimate PU from illegal SU in the physical layer will be designed for that both such PU and SU share the same spectrum. The specific innovations of the thesis are as follows:In the first part, two schemes are proposed:TDMA and TDMA combined with SDMA. Both schemes apply receiving antenna selection to reduce the complexity, and select the SU based on the trace to maxmize the throughput of cognitive network, with limited uplink feedback. The first solution selects single SU in a time slot. The second scheme selects multi-SUs, which apply the best antenna with precoding for communication. In addition, this scheme considers SU channel the interference to the PU channel. Its feedback is a vector. From the simulation result, in the field of cognitive network throughput, the first proposed scheme can approach the scheme that CBS knows the knowledge of CSIT. The second scheme improved a lot, because more SUs are scheduled in one time slot. However, the average rate is lower than the one in the first scheme.In the second part, the security threat is considered for the risk of the same spectrum sharing. Based on the physical layer, we propose a scheme which combines the acquisition of channel state information with artificial noise scrambled. It guarantees the legitimate PU obtaining the near-accurate channel information, and artificial noise can be cancelled while the illegal SU gets its channel information with huge error instead. Furthermore, such SU is not able to decode the artificial noise, what makes BER performance worse. The simulation result shows:in one hand, legitimate PU can decode the message with low BER, with the illegal SU still decoding them with the performance BER≈0.5. In the other hand, our proposed scheme has better performance over the artificial noise scrambled scheme only in the field of secure capacity.