The presence of compressed sensing theory makes the sample rate relate to the signal structure and content. The
sample rate of compressed sensing for the signal sample, coding and reconstruction is less than the Nyquist theorem methods.
Therefore, a method is proposed to solve the bottleneck problem of data redundancy and resource-wasting. Moreover, it offers
new developing chances for other research fields. The development and current situation of compressed sensing theory are
involved. A detailed carding and research on the sparse representation, measurement matrix design, reconstruction algorithm
and application aspects is discussed. Therefore, the current hot spots and difficulties are analyzed and discussed. Finally, the
direction of future development and application prospect are discussed.