High Spectrum and Efficiency Improved Structured Compressive Sensing-Based Channel Estimation Scheme for Massive MIMO Systems

Abstract

AbstractDue to its high spectrum and energy proficiency, massive MIMO will become the most promising technique for 5G communications in future. For accurate channel estimation, potential performance gain is essential. The pilot overhead in conventional channel approximation schemes is due to the enormous number of antennas used at the base station (BS), and also this will be too expensive; for frequency division duplex (FDD) massive MIMO, it is very much unaffordable. We introduced a structured compressive sensing (SCS)-based temporal joint channel estimation scheme which reduces pilot overhead where it requires, delay-domain MIMO channels are leveraged whereby the spatiotemporal common sparsity. The accurate channel estimation is required to fully exploit the mass array gain, which states the information at the transmitter side. However, FDD downlink channel estimation always requires more training and computation than TDD mode, even though the uplink and downlink channel is always not straightforwardly reciprocal, due to the massive number of antennas in base station. At the base station, we first introduce the non-orthogonal pilots which come under the structure of compressive sensing theory to reduce the pilot overhead where they are required. Then, a structured compressive sensing (SCS) algorithm is introduced to approximate the channels associated with all the other OFDM symbols in multiple forms, then the inadequate number of pilots is estimated, and the spatiotemporal common sparsity of massive MIMO channels is also exploited to recover the channel estimation with precision. Furthermore, we recommend a space–time adaptive pilot scheme to decrease the pilot overhead, by making use of the spatiotemporal channel correlation. Additionally, in the multi-cell scenario, we discussed the proposed channel estimation scheme. The spatial correlation in the wireless channels is exploited for outdoor communication scenarios, where mostly in wireless channels. Meanwhile, compared with the long signal transmission distance, the scale of the transmit antenna is negligible. By utilizing the greater number of spatial freedoms in massive MIMO can rise the system capacity and energy proficiency of magnitude. Simulation results will show that the proposed system outperforms than all the other existing systems.KeywordsFrequency division duplexingMassive MIMOStructured compressive sensingPilot overhead