Frame Synchronization Technology Research Based On Compressed Sampling

The Frame Timing Synchronization(FTS)technology has been become the hot topic of the communication and its performance directly affects the communication quality of entire system.In the traditional signal processing,the frame synchronization needs to obey the Nyquist theorem.However,a mass of sampled data greatly increases sampling devices' requirement,and then increases hardware cost and energy consumption of the system.Since the sparsity or compressibility property of signal,the signals can be sampled by the technology of Compressed Sampling(CS)at far below the Nyquist sampling rate without loss of information.This paper further researches the CS theory which is applicated in FTS.Instead of preserving the measurement amplitudes of CS-based technology,the single-bit CS technology only preserves measurements' symbol information,so single-bit CS technology can further reduce the system's energy consumption and the Analog-to-Digital Converter(ADC)'s design difficulty.In view of the above-mentioned advantages,the single-bit CS technology is introduced into FTS in this paper.Firstly,the proposed method performs single-bit compressive sampling on the received signal in the frame timing transformation domain.Subsequently,the timing metrics of frame timing synchronization are reconstructed by using the sampled bit stream.Finally,the index position of the FTS is found by searching the reconstructed timing metrics according to the correlation criterion.Compared with the CS-based FTS method,the analysis and simulation results show that the proposed method can improve the correct synchronization probability(CSP)with the same bit overhead.Correspondingly,less bit overhead is needed in the proposed method when the CSP is same.Meanwhile,the proposed method only requires a comparator in quantization phase,which reduces the design difficulty of ADC.Then this paper researches FTS method based the traditional Superimposed Training Sequence(STS).In view of the redundant sampled data during transmission,the CS method is introduced into FTC based STS in this paper.Firstly,the training sequence is weighted superimposed on the modulated sparse signal.Subsequently,the proposed method performs compressive sampling on the received signal.Then,the reconstructed signals are reconstructed by using the sampled bit stream.The index position of the FTS is found by searching the reconstructed signals.Finally,the TST signals are subtracted from the reconstructed signals and the original sparse signals are recovered by demodulation.The proposed method can reduce the sampled data,thus the ADC's accuracy requirement is greatly decreased.Compared with the traditional STS-based FTS method,the analysis and simulation results show that the proposed method can achieve approximately equivalent CSP in the condition of high Signal-to-Noise Ratio(SNR).Meanwhile,the proposed method is robust under different parameters.Nowadays,the CS theory is widely used in electronic engineering,computer science,wireless communication,statistics,Applied Mathematics and other fields,and a lot of research results have been achieved,such as speech recognition,intellignet house,connected vehicles,automated factories,Internet of Things,radar detection,medical imaging,image acquisition and processing,channel estimation,face recognition.