Efficient downlink channel reconstruction for FDD transmission systems

Abstract

Abstract-In this paper, a novel scheme is proposed to reconstruct downlink channel for frequency-division-duplex (FDD) systems from uplink channel state information (CSI) with limited feedback. We adopt the Newtonized orthogonal matching pursuit (NOMP) algorithm to estimate the frequency-independent parameters of the delays of a multi-path channel through uplink training. The gain of each path is further refined through downlink training using least squares (LS) algorithm and sent back to the base station (BS). Using the uplink-estimated delays and the downlink-refined gains, the BS reconstructs the downlink multipath channel. Simulations and over-the-air tests are conducted to examine the performance of the proposed NOMP-LSbased reconstruction scheme for the downlink channel. Results confirm the effectiveness of the NOMP-LS scheme and that the reconstructed channel is accurate.

FDD Massive MIMO Based on Efficient Downlink Channel Reconstruction

“…These paths are projected on different optimal grid points but share a common suboptimal grid point. According to (23), if the BS beamforms the downlink pilots only on the direction of this common suboptimal grid point, then the coefficient matrix A becomes…”

“…However, [22] did not consider delays, set restriction on the estimated number of paths, and only retained four dominant paths. [23]- [25] merely provided a channel reconstruction scheme to obtain the downlink channel of a single user. Given that [23]- [25] used user-dedicated pilots for downlink training, if we simply extend the method to fit multiuser scenarios, then a large amount of training resource is needed to estimate downlink gains of multiuser channels, as designed in [22].…”

“…[23]- [25] merely provided a channel reconstruction scheme to obtain the downlink channel of a single user. Given that [23]- [25] used user-dedicated pilots for downlink training, if we simply extend the method to fit multiuser scenarios, then a large amount of training resource is needed to estimate downlink gains of multiuser channels, as designed in [22]. Moreover, the schemes proposed in [23] and [24] are inapplicable in massive 3D-MIMO systems because they can not extract the full-dimensional spatial parameters.…”

A Covariance-Based Hybrid Channel Feedback in FDD Massive MIMO Systems

“…To solve this well recognized problem, a large array of strategies have been proposed to reduce FDD-based MIMO feedback overhead, including recent efforts tackling FDD-based massive MIMO specifically, such as advanced trellis-extended codebook design [8], compressive sensing-based channel feedback reduction [9], [10], antenna grouping-based feedback reduction technique [11], frequency-independent parameter extraction and downlink channel reconstruction [12], [13], angular domain energy distribution-based channel estimation [14], [15] etc. Moreover, the limited channel feedback issue was also tackled by exploiting user cooperation via device-to-device communications [16], [17].…”

Downlink Channel Tracking for FDD Large-Scale Antenna Systems

“…Due to the large amount of pilots and feedback, nonparametric methods are hardly applied in frequency-division-duplex (FDD) large scale antanna systemS (LSASs), while parametric methods tackle this bottleneck. [6] and [7] utilize the Newtonized orthogonal matching pursuit (NOMP) method to extract parameters of each path in uplink and take advantage of the frequency-independent parameters obtained from the uplink to reconstruct the downlink channel. Another method for FDD three-dimensional (3D) Multi-Antenna Systems has been proposed in [8], which employs uniform planar array (UPA) at the BS and can estimate the downtilt and the azimuth.…”