Practical beamforming based on RSSI measurements using off-the-shelf wireless clients

Abstract

WLANs have become an important last-mile technology for providing internet access within homes and enterprises. In such indoor deployments, the wireless channel suffers from significant multipath scattering and fading that degrades performance. Beamforming is a smart antenna technology that adjusts the transmissions at the transmitter to reenforce the signals received through multiple paths at the receiver. However, doing this requires the accurate estimation of the channel coefficients at the receiver and its knowledge at the transmitter which off-the-shelf WiFi clients are incapable of doing. In this work, we develop a novel procedure that uses Received Signal Strength Indicator (RSSI) measurements at the receiver along with an intelligent estimation methodology at the transmitter to achieve beamforming benefits. Using experiments in an indoor office scenario with commercial WiFi clients, we show that the scheme achieves significant performance improvements across diverse scenarios.

FastBeam: Practical fast beamforming for indoor environments

“…We implement F astBeam on a Phocus Array System ( [3]) with eight antennas and use experimental evaluations to study its performance in indoor environments. We show that the time complexity for RSSI based beamforming can be reduced by an average of 50% and by as much as 75% compared to existing approaches ( [2]). Such a reduction in the time complexity has a significant impact on the throughput performance under several conditions: we show that the throughput performance of F astBeam compared to the existing work is 1.4× better on average, and up to 1.8× better in the best case.…”

“…Specifically, in indoor environments, while simple directional antenna solutions suffer from multipath fading, adaptive beamforming still holds tremendous promise. A recent advancement in beamforming is a partial CSI approach called RSSI based beamforming [2], which only uses RSSI measurements sent as feedback from the client to make beamforming decisions. While it is provably optimal (same performance as traditional beamforming under static channel conditions), the most attractive property of RSSI based beamforming is that it does not require the receiver to be equipped with an antenna array or special hardware to provide benefits.…”

“…(Actually, since the exact value of cos θ can be known, 2 sets of tandem activated antennas is sufficient. However, since time-complexity is not a concern in [2], it does not deal with this issue. In section IV, we will provide an even more efficient algorithm using only 1 set of tandem activated antennas.…”

“…However, collecting such information at the receiver requires specialized hardware and complex processing. Recent techniques [2] have demonstrated that optimal beamforming can be achieved while relying purely on the receive signal strength indicator (RSSI), a parameter that is readily and easily available on most wireless cards.…”

“…However, they either use directional antennas with fixed radiation patterns ( [4], [5], and [6]) that are less sophisticated than beamforming, or require the usage of CSI ( [7]) that require the clients to be equipped with special hardware. In [2], an algorithm is presented to enable an off-the-shelf client to gain beamforming benefits without hardware modifications. However, we show that the proposed algorithm (called SimpleBeam in this paper) is not suitable for time varying channels.…”