Vision-Aided 6G Wireless Communications: Blockage Prediction and Proactive Handoff

Abstract

The sensitivity to blockages is a key challenge for the high-frequency (5G millimeter wave and 6G sub-terahertz) wireless networks. Since these networks mainly rely on line-of-sight (LOS) links, sudden link blockages highly threaten the reliability of the networks. Further, when the LOS link is blocked, the network typically needs to hand off the user to another LOS basestation, which may incur critical time latency, especially if a search over a large codebook of narrow beams is needed. A promising way to tackle the reliability and latency challenges lies in enabling proaction in wireless networks. Proaction basically allows the network to anticipate blockages, especially dynamic blockages, and initiate user hand-off beforehand. This paper presents a complete machine learning framework for enabling proaction in wireless networks relying on visual data captured, for example, by RGB cameras deployed at the base stations. In particular, the paper proposes a vision-aided wireless communication solution that utilizes bimodal machine learning to perform proactive blockage prediction and user hand-off. The bedrock of this solution is a deep learning algorithm that learns from visual and wireless data how to predict incoming blockages. The predictions of this algorithm are used by the wireless network to proactively initiate hand-off decisions and avoid any unnecessary latency. The algorithm is developed on a vision-wireless dataset generated using the ViWi data-generation framework. Experimental results on two basestations with different cameras indicate that the algorithm is capable of accurately detecting incoming blockages more than ∼ 90% of the time. Such blockage prediction ability is directly reflected in the accuracy of proactive hand-off, which also approaches 87%. This highlights a promising direction for enabling high reliability and low latency in future wireless networks.

LiDAR-Aided Mobile Blockage Prediction in Real-World Millimeter Wave Systems

“…To draw some insights about the potential gains of the proposed methods for the initial access latency in 3GPP 5G NR, we adopt the approach in [8] for analyzing the results. According to the 3GPP specifications, a conventional reactive hand-off results in an overall delay of 222.8ms.…”

“…According to the 3GPP specifications, a conventional reactive hand-off results in an overall delay of 222.8ms. If the blockage is predicted proactively, a successful proactive hand-off scenario results in 11.4ms latency associated with the contention free random access [8]. Based on the prediction accuracy and the latency given by the average latency for the user hand-off, the average latency δ for the user is given by δ = p × 11.4 + (1 − p) × 222.8, where, p is the blockage prediction accuracy at 0.1s, 0.5s and 1s.…”

“…The approaches in [6], [7], are mainly suitable for predicting stationary blockages or blockages that are a few milliseconds ahead. To predict future blockages hundreds of milliseconds before they happen, leveraging other sensory data such as cameras [8]- [11] was proposed. Deploying cameras, however, may not always be possible for privacy/regulatory reasons.…”

Prediction of mmWave/THz Link Blockages through Meta-Learning and Recurrent Neural Networks

“…The use of Recurrent Neural Networks (RNNs) to predict beam blockages in scenarios with user mobility has been studied in [7]- [10], which learn the spatial and/or temporal correlation of blockages. Furthermore, visual features has been shown to increase the accuracy of such predictors [5], [11]. In [12], the authors use the dynamics of the signal-to-noise ratio (SNR) prior to a blockage, such as diffraction effects, to predict blockages.…”

“…Our goal is to predict future blockages between the BS and a device indexed by k based on previous observations of the channel and, possibly, on side information. The sequence of observations is denoted o 1:t , where each observation vector o t may contain, among others, SNR estimates, device locations, and images captured by cameras [5], [11], at time steps 1, 2, . .…”