Haifan Yin

Decontaminating pilots in massive MIMO systems

Pilot contamination is known to severely limit the performance of large-scale antenna (“massive MIMO”) systems due to degraded channel estimation. This paper proposes a twofold approach to this problem. First we show analytically that pilot contamination can be made to vanish asymptotically in the number of antennas for a certain class of channel fading statistics. The key lies in setting a suitable condition on the second order statistics for desired and interference signals. Second we show how a coordinated user-to-pilot assignment method can be devised to help fulfill this condition in practical networks. Large gains are illustrated in our simulations for even small antenna array sizes.

Robust Pilot Decontamination Based on Joint Angle and Power Domain Discrimination

We address the problem of noise and interference corrupted channel estimation in massive MIMO systems. Interference, which originates from pilot reuse (or contamination), can in principle be discriminated on the basis of the distributions of path angles and amplitudes. In this paper, we propose novel robust channel estimation algorithms exploiting path diversity in both angle and power domains, relying on a suitable combination of the spatial filtering and amplitude based projection. The proposed approaches are able to cope with a wide range of system and topology scenarios, including those where, unlike in previous works, interference channel may overlap with desired channels in terms of multipath angles of arrival or exceed them in terms of received power. In particular, we establish analytically the conditions under which the proposed channel estimator is fully decontaminated. Simulation results confirm the overall system gains when using the new methods.

Enabling massive MIMO systems in the FDD mode thanks to D2D communications

We propose novel approaches to design feedback in FDD massive MIMO systems. We exploit synergies between massive MIMO systems and inter-user communications based on D2D. The exchange of local CSI among users, enabled by D2D communications, makes available global CSI at the terminals. Thus, we can construct more informative forms of feedback based on this shared knowledge. Two feedback variants are highlighted: 1) cooperative CSI feedback, and 2) cooperative precoder index feedback. For a given feedback overhead, the sum-rate performance is assessed and the gains compared with a conventional massive MIMO setup without D2D are shown.

Pilot decontamination using combined angular and amplitude based projections in massive MIMO systems

We address the problem of noise and interference corrupted channel estimation in massive MIMO systems. Interference, which originates from pilot reuse (or contamination), can in principle be discriminated from the desired channels upon observing the distributions of path angles and amplitudes. In this paper we propose novel robust channel estimation algorithms exploiting path diversity in both angle and amplitude domains, relying on a suitable combination of a subspace projection and MMSE estimation. The proposed estimator improves on past methods in a wide range of system and topology scenarios.

Precoder feedback versus channel feedback in massive MIMO under user cooperation

In multiuser massive MIMO systems, it is not clear whether users should feed back the channel or the precoder when they can exchange the channel state information (CSI). This paper compares the precoder feedback scheme with the channel feedback scheme. It is found that when there are sufficient number of bits for CSI exchange, the precoder feedback scheme can reduce the interference leakage to 1 /(K - 1) of the channel feedback scheme, where K is the number of users. Moreover, the interference leakage under the precoder feedback scheme decreases faster than the channel feedback scheme when the number of feedback bits increases.

Feedback Mechanisms for FDD Massive MIMO With D2D-Based Limited CSI Sharing

Channel state information (CSI) feedback is a challenging issue in frequency division duplexing (FDD) massive MIMO systems. This paper studies a cooperative feedback scheme, where the users first exchange their CSI with each other through device-to-device (D2D) communications, then compute the precoder by themselves, and feedback the precoder to the base station (BS). Analytical results are derived to show that the cooperative precoder feedback is more efficient than the CSI feedback in terms of interference mitigation. To reduce the delays for CSI exchange, we develop an adaptive CSI exchange strategy based on signal subspace projection and optimal bit partition. Numerical results demonstrate that the proposed cooperative precoder feedback scheme with adaptive CSI exchange significantly outperforms the CSI feedback scheme, even under moderate delays for CSI exchange via D2D.

Dual-Regularized Feedback and Precoding for D2D-Assisted MIMO Systems

This paper considers the problem of efficient feedback design for massive multiple-input multiple-output (MIMO) downlink transmissions in frequency division duplexing (FDD) bands, where some partial channel state information (CSI) can be directly exchanged between users via device-to-device (D2D) communications. Drawing inspiration from classical point-to-point MIMO, where efficient mechanisms are obtained by feeding back directly the precoder, this paper proposes a new approach to bridge the channel feedback and the precoder feedback by the joint design of the feedback and precoding strategy following a team decision framework. Specifically, the users and the base station (BS) minimize a common mean squared error (MSE) metric based on their individual observations on the imperfect global CSI. The solutions are found to take similar forms as the regularized zero-forcing (RZF) precoder, with additional regularizations that capture any level of uncertainty in the exchanged CSI, in case the D2D links are absent or unreliable. Numerical results demonstrate superior performance of the proposed scheme for an arbitrary D2D link quality setup.

A statistical approach to interference reduction in distributed large-scale antenna systems

This paper considers the problem of interference control in networks where base stations signals are coherently combined (aka network MIMO). Building on an analogy with so-called massive MIMO, we show how second-order statistical properties of channels can be exploited when the massive MIMO array corresponds in fact to many antennas randomly spread over a two-dimensional network. Based on the classical one-ring model, we characterize the low-rankness of channel covariance matrices and show the rank is related to the scattering radius. The application of the low-rankness property to channel estimations denoising and low complexity interference filtering is highlighted.

Dual-regularized precoding: A robust approach for D2D-enabled massive MIMO

This paper designs efficient feedback mechanisms to help enable massive MIMO in frequency division multiplexing (FDD) bands. By exploiting possible device-to-device (D2D) coordination and using a team decision approach, a scheme is developed to bridge the feedback in the channel space and that in the precoder space. It is found that the desired feedback and precoding vectors take similar forms as the regularized zero-forcing (RZF) precoder, with additional regularizations to capture any level of uncertainty of the exchanged channel state information (CSI). Numerical results demonstrate superior performance of the proposed scheme for an arbitrary D2D link quality setup.

Enhancing LTE with Cloud-RAN and Load-Controlled Parasitic Antenna Arrays

Cloud radio access network systems, consisting of remote radio heads densely distributed in a coverage area and connected by optical fibers to a cloud infrastructure with large computational capabilities, have the potential to meet the ambitious objectives of next generation mobile networks. Actual implementations of C-RANs tackle fundamental technical and economic challenges. In this article, we present an end-to-end solution for practically implementable C-RANs by providing innovative solutions to key issues such as the design of cost-effective hardware and power-effective signals for RRHs, efficient design and distribution of data and control traffic for coordinated communications, and conception of a flexible and elastic architecture supporting dynamic allocation of both the densely distributed RRHs and the centralized processing resources in the cloud to create virtual base stations. More specifically, we propose a novel antenna array architecture called load-controlled parasitic antenna array (LCPAA) where multiple antennas are fed by a single RF chain. Energy- and spectral-efficient modulation as well as signaling schemes that are easy to implement are also provided. Additionally, the design presented for the fronthaul enables flexibility and elasticity in resource allocation to support BS virtualization. A layered design of information control for the proposed end-to-end solution is presented. The feasibility and effectiveness of such an LCPAA-enabled C-RAN system setup has been validated through an over-the-air demonstration.