Research On Singal Detection Based On Compress Sensing

With the rapid development of information technology, the IMT-Advanced system of the fourth generation (4G) mobile communication system is being widely concerned. As the key technology of the fourth generation mobile communication system, orthogonal frequency division multiplexing (OFDM) has been studied extensively. OFDM technology mainly has two advantages. First it has the quantity of high spectrum utilization. Second it has a good effect on resisting the multipath fading; So OFDM is chosen as the key technology of physical layer of4G communication system. In the field of signal recognition, blind identification technology is always a hotspot; blind detection of OFDM has a significant role for the military. Therefore, from the national security considerations, the study is of great significance. On the other hand, with the increasing signal bandwidth, the demands to the sampling rate and processing speed is increasingly higher. Limited by Nyquist theory, the improvement of signal sampling rate has encountered bottleneck. In2006Candes and Donoho proposed the concept of compressive sensing, the concept completely broke the traditional sampling frequency theory that the sampling frequency should be two times greater than the bandwidth of the system, and it is of great significance to reducing the sampling rate.Considering the compressive sensing technology and blind detection of OFDM signal, the main contents of this thesis are as follows:First, this thesis introduces the key technologies of OFDM signal, and establishes the mathematical model of OFDM signals and SCLD signals, then it deduce the second order cyclic cumulates of OFDM signals and SCLD signals. The simulation shows that the second order cyclic cumulate of OFDM and SCLDs signals is sparse in time and frequency. Utilizing the feature of signal in time and frequency domain, we proposed a new measure to detect signals. At the same time, we take advantage of the quantity of the non-zero point; we can get some parameters of OFDM signals. Simulation shows that the proposed method performs better than traditional method.Second, this thesis has research on the compressive sensing theory and its background. The compressive sensing theory mainly contains three key techniques:sparse matrix, observation matrix and recovery algorithm. By utilizing the characteristic of the OFDM cyclic cumulates, we propose a new method on the OFDM signal blind detection based on compressive sensing which combines the blind identification and compressive sensing technique effectively.Third, in the compressed signal mapping processing, there will be a large matrix. So it has very high request to the storage. To solve the big matrix problem, this thesis proposes a new method to calculate the cyclic cumulates of OFDM signals. The central idea of this method is reducing the storage space by the cycle accumulation method.Forth, in simulation, the results showed the correctness of the method. This method can recover cyclic cumulates of OFDM signals effectively. The recovered cyclic cumulates of signals can be used to blind detection and parameters detection correctly. It shows that the method has a good performance even in low signal to noise ratio (SNR).