Research On Compression Sampling Mechanism And Methods

With the increasing of wireless communication rate,the bandwidth becomes wider and wider.According to the Nyquist theorem,ADCs’ sampling rate must be,at least,twice the bandwidth to guarantee the signal accurate recovery.It is a big challenge for analog to digital convertors(ADCs).Thus the higher rate sampling ADCs becomes a bottleneck,which hinders the higher rate communication.A new theory called compressed sampling(CS)indicates that an analog sparse signal can be recovered from fewer numbers of samples than the least sampling number stated in the Nyquist theorem.Thus,it has become a hot research topic.This paper mainly studies the low complexcity sampling structure in compressed sampling.In practical implementations,analog-to-information conversion(AIC)is used to obtain compressed samples directly from an analog signal.Several sampling structures of AIC has been proposed,such as random demodulation(RD),modulated wideband converter(MWC),among which RD and MWC are the most common structures.However,RD fails in reconstructing multiband signals which limits its application.Moreover,though MWC can be applied into the compressed sampling of multiband signals,it has high implementation complexity owing to the fact that its measurement matrix is dense.Therefore,finding a reliable CS sampling structure which has low implementation complexity and can be applied to a wide band signal becomes a new research topic.To reduce the complexity of MWC,this paper proposes the diagonal modulated wideband converter(D-MWC),where each BMI only works within a partial time period that is non-overlapping instead of each BMI working all the time in MWC.The measurement matrix of D-MWC is sparse that can significantly simplify the sampling structure.Simulations verify the effectiveness of D-MWC.The proposed D-MWC offers a good tradeoff between complexity and sampling performance.Furthermore,this paper proposes a more general sampling structure n-DMWC based on D-MWC,where n is a positive integer.It has the same structure as D-MWC but the working time of each BMI during a sampling period T is different.We suppose there are K branches and the sampling rate is 1/T both in MWC and nD-MWC.The working time of each BMI in nD-MWC is nT/K instead of T in D-MWC during a sampling period T.Compared with D-MWC,nD-MWC not only has a similar low complexity,but also can get better performance by changing the parameters of N,which has better flexibility.Simulations verify the effectiveness of nD-MWC.Therefore,the proposed nD-MWC is a good AIC structures because of low complexity and similar performance,and it has better generality compared to D-MWC.