Analysis And Implementation Of Frequency Hopping Signals Based On Compressed Sensing

The carrier frequency of frequency hopping signals can hop in accordance with a pseudo random sequence with certain regulations.FH has many advantages,such as antiinterference,easy networking and low interception.Therefore,it has been widely applied in military and civil fields.Traditional analysis methods for frequency hopping signals are based on Nyquist sampling theorem.But with the continuous development of communication technology and the increasing bandwidth of frequency hopping signals,the analysis and processing of frequency hopping signals have encountered bottlenecks.At the same time,the compressed sensing theory for sparse signal has been born.It can be used to sample with far below the sampling rate of Nyquist,and it can accurately reconstruct the original signal,which opens a new door for the analysis and processing of the frequency hopping signals.Therefore,this paper mainly analyzes the frequency hopping signal analysis based on the compressed sensing theory,and designs and implements frequency hopping signal analysis system with compressed sensing at its core.The main research works are listed as follows:1.The basic principles of frequency hopping communication and compressed sensing are introduced.Several common observation matrices are compared,and the success probability of signal reconstruction with different number of observations and sparsity is demonstrated and analyzed.2.The data acquisition scheme of compressed sensing is studied.The AIC structure for analog sparse signals and the MWC structure for multiband signals are carefully analyzed,due to hardware limitations and the large number of required channels in the sampling process,a new scheme is proposed to reduce the number of subchannels at the cost of the falling sampling rate of subchannels and other digital processing,digital channelized based on polyphase filter structure is introduced.3.The signal reconstruction algorithm of compressed sensing theory is studied.Based on MP and OMP algorithm respectively,MPMMV and OMPMMV algorithms for MMV problems are proposed.Considering the feasibility of engineering implementation,the former has more advantages in hardware implementation.In view of the selection of initial residuals in the reconstruction process,the feasibility of taking the observation value matrix as the initial residuals is demonstrated,and the simplification of the construction structure is realized.4.Frequency hopping analysis system based on compressed sensing theory is designed and implemented.This paper completed the configuration of the transmitter of the high-speed transceivers for sending a pseudo-random sequence,and implemented the multi-phase filtering channelization structure and the reconstruction process of compressed sensing on FPGA,and debugged and verified it with the hardware platform of the project.