Compressed sensing (CS) is an emerging field concerning the sampling and compression of signals.In practice, by choosing a proper basis, many signals have sparse representations whose nonzero coefficients are very few.Essentially, CS provides a fundamentally new approach to deal with sparse signals.It combines the sampling and compression of signals by merely taking some nonadaptive linear projections.While it recovers sparse signal by solving a optimization problem.Comparing with traditional sampling method, CS can recover a sparse signal from surprisingly few measurements.Because of its elegancy in theoretic framework and practical implementation, CS captures a tremendous amount of interest from both mathematical and engineering communities.