Dynamic measurement for compressed sensing based channel estimation in OFDM systems

In most of existing compressed sensing based channel estimation method for OFDM systems, channel impulse response (CIR) is treated as finite-length vector and involved overall for the measuring and reconstruction, which induces a delay proportional to vector size to obtaining channel estimates. In this paper, a dynamic measurement scheme is proposed for compressed sensing of OFDM channel to reduce estimation delay. The CIR corresponding to successive OFDM symbols is considered as continuous-time sparse signal streams, and real-time measurements of channel are obtained by employing dimension-fixed and overlapped measurement matrices which have the block-diagonal structure. The overlapping portions of successive measurements is exploiting to decrease the computational complexity of measurement process. Theoretical and simulation results show that our dynamic compressed sensing measurement can significantly reduce the reconstruction delay with the similar estimation accuracy to that by traditional random compressed sensing matrices, which verifies the effectiveness of the proposed method.