Compressive wireless data transmissions under channel perturbation

Compressed sensing (CS) technique has attracted a lot of recent research interests in mathematics and signal processing fields. Literature studies often exploit CS at the receiver side to sub-sample the receiving signals to reduce the sampling rate and processing overhead. It would be of great benefit if it is possible to exploit CS at the transmitter side to reduce the redundancy of the data before transmission to conserve precious wireless bandwidth. Different from receiver-side sub-sampling, the sub-sampled transmitting data may be perturbed by the dynamics of wireless channels and experience higher overall noise. In this paper, we propose a set of mechanisms to enable compressive wireless data transmissions. Specifically, we investigate the impacts of imperfect channel equalization on the data reconstruction, and propose a comprehensive signal recovery algorithm to cope with the perturbations introduced by wireless channels. Simulation results demonstrate that our proposed schemes can effectively reduce the effects of dynamic wireless channels on the data reconstruction and maintain the performance comparable to that of traditional communication scheme which does not apply CS to compress data. This indicates that it is promising to exploit CS to reduce the communication data thus bandwidth requirement. Transmission data reduction can complement existing efforts of improving wireless channel capacity to support the quick growth of wireless applications.