Adaptive Blind Beamforming for Intelligent Surface

TL;DR

This paper proposes a grouping strategy that enables adaptive blind beamforming and validates the advantage of the proposed blind beamforming algorithm in the real-world networks at 3.5 GHz aside from simulations.

Abstract

Configuring intelligent surface (IS) or passive antenna array without any channel knowledge, namely blind beamforming, is a frontier research topic in the wireless communication field. Existing methods in the previous literature for blind beamforming include the RFocus and the CSM, the effectiveness of which has been demonstrated on hardware prototypes. However, this paper points out a subtle issue with these blind beamforming algorithms: the RFocus and the CSM may fail to work in the non-line-of-sight (NLoS) channel case. To address this issue, we suggest a grouping strategy that enables adaptive blind beamforming. Specifically, the reflective elements (REs) of the IS are divided into three groups; each group is configured randomly to obtain a dataset of random samples. We then extract the statistical feature of the wireless environment from the random samples, thereby coordinating phase shifts of the IS without channel acquisition. The RE grouping plays a critical role in guaranteeing performance gain in the NLoS case. In particular, if we place all the REs in the same group, the proposed algorithm would reduce to the RFocus and the CSM. We validate the advantage of the proposed blind beamforming algorithm in the real-world networks at 3.5 GHz aside from simulations.

Figures (18)

• Figure 1: Line-of-sight (LoS) case vs. non-line-of--sight (NLoS) case for the IS-assisted transmission. As a subtle issue discovered in this work, the existing blind beamforming algorithms, the RFocus Arun2020RFocus and the CSM ren2022configuring, can fail to work in the NLoS case.
• Figure 2: An example of the alternating CSM algorithm with the REs divided into two groups. Assume that $N=4$ and the channels are $[h_0,h_1,h_2,h_3,h_4]=[0,1.7646+2.1012j,0.2792-1.6644j,0.7178+3.1842j,0.6117-2.2282j]\times10^{-5}$; the reflected channels are grouped as $S_{\mathrm{I}}=\{h_1, h_4\}$ and $S^c_{\mathrm{I}}=\{h_2, h_3\}$. We assume that $T\rightarrow\infty$ and thus each empirical conditional average has converged to the actual conditional expectation.
• Figure 3: Visualization of the procedure of Algorithm \ref{['alg:GCSM']}.
• Figure 4: Our field test uses an IS that comprises 400 REs and provides 4 phase shift options $\{0,\pi/2,\pi,3\pi/2\}$ for each RE. For the LoS case, omnidirectional antennas are deployed at all devices; for the NLoS case, directional antennas are used to prevent direct signal propagation from the transmitter to each receiver.
• Figure 5: The IS-assisted downlink network considered in our simulations. One or more receivers are randomly located within the shaded area.

Theorems & Definitions (11)

• Lemma 1
• Example 1
• Example 2
• Remark 1
• Remark 2
• Remark 3: Preview of Why RFocus and CSM may Fail
• Proposition 1
• Proposition 2
• Remark 4
• Proposition 3