Research On Timing Synchronization Sequence Sparse Reconstruction And Frame Synchronization Algorithm

Multiple Input Multiple Output-Orthogonal Frequency DivisionMultiplexing (MIMO-OFDM) technology can greatly improve systemcapacity and spectrum efficiency combating with multipath interference. It isone of the key technologies of4G. Although MIMO-OFDM technology has somany advantages, it is still very sensitive for timing synchronizationperformance. So the timing synchronization research has been a researchhotspot of wireless communications.In recent years, the Tao and Candes et al. proposed a new informationsampling theory called Compressed Sensing (CS). The theory is different fromthe traditional Nyquist sampling theorem. It mentioned: If the signal has asparse nature, CS can be used to obtain its internal information. Thisnon-linear sampling rate will be much lower than the Nyquist sampling rate,and finally the original signal can be accurately recovered by appropriatereconstruction method. CS theory is used to solve the traditional technicaldifficulties and becomes a new research hotspot. There are hardly any existingresearch about introducing compressed sensing theory to the timingsynchronization, so the applications of compressed sensing theory in timingsynchronization is a worth research direction.In this paper, we will exploratory research compressed sensing whichapplications in timing synchronization, mainly to study the framesynchronization problems. Because of timing synchronization sequences usedcomplex sequence that different with real signal in common. So, first studywill research the synchronization sequence sparse reconstruction using theparticularity of synchronization sequence. Under different observation matrixand different observation lengths, we have realized the synchronizationsequence sparse reconstruction. Simulation results show that synchronizationsequence after sparse can achieve a good reconstruction, and also havecompared three common observation matrix performance differences in theexperiment.Then this paper research the problem of in the traditional timingsynchronization algorithms for distributed MIMO-OFDM, the robustness ofmetric value is poor while changing the number of transmitting antennas. And whether receivers capture the preamble is determined by the threshold withexperience value. By theory analysis, when the number of transmittingantenna changes, the traditional synchronization metric function with thetraditional selection method for threshold will greatly increase the probabilityof miss capture and mis-capture. In order to solve this problem, afteranalyzing the statistical characteristics of metric function, we modified thetiming synchronization metric function and deduced its statistical properties.And then, a frame synchronization algorithm based on computing threshold(CTFS) is proposed using the‘3σ’principle. At the last, the CTFS framesynchronization algorithm is compared with the traditional UPSP and OVPSPalgorithm by Monte-Carlo simulation in the different systems with differentMIMO transmitting antenna numbers. The numerous simulations proved thatthe CTFS algorithm has greatly improved the synchronization performanceand robustness of synchronization metric value. This work has basicallysolved the serious problem in traditional synchronization function when thenumber of transmit antennas changes.