Beam Training and Data Transmission Optimization in Millimeter-Wave Vehicular Networks

Scalabrin, Maria; Michelusi, Nicolò; Rossi, Michele

doi:10.1109/glocom.2018.8647890

Cited by 16 publications

(14 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…3, we show the training progress of DR-VAE in terms of average ELBO, KL divergence, spectral efficiency and BT overhead (percentage of the frame duration used for BT). We compare different frameworks to learn the transition model: the proposed DR-VAE, the naive approach (see (6)) and the Baum-Welch algorithm [11]. Observations the decoupling of policy design and mobility learning via the dual-timescale approach: we can thus train multiple mobility models simultaneously and use the prior belief generated based on any one of the mobility models.…”

Section: Simulation Setupmentioning

confidence: 99%

Learning and Adaptation for Millimeter-Wave Beam Tracking and Training: a Dual Timescale Variational Framework

Hussain¹,

Michelusi²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Millimeter-wave vehicular networks incur enormous beam-training overhead to enable narrow-beam communications. This paper proposes a learning and adaptation framework in which the dynamics of the communication beams are learned and then exploited to design adaptive beam-training with low overhead: on a long-timescale, a deep recurrent variational autoencoder (DR-VAE) uses noisy beamtraining observations to learn a probabilistic model of beam dynamics; on a short-timescale, an adaptive beam-training procedure is formulated as a partially observable (PO-) Markov decision process (MDP) and optimized via point-based value iteration (PBVI) by leveraging beam-training feedback and a probabilistic prediction of the strongest beam pair provided by the DR-VAE. In turn, beam-training observations are used to refine the DR-VAE via stochastic gradient ascent in a continuous process of learning and adaptation. The proposed DR-VAE mobility learning framework learns accurate beam dynamics: it reduces the Kullback-Leibler divergence between the ground truth and the learned beam dynamics model by 86% over the Baum-Welch algorithm and by 92% over a naive mobility learningapproach that neglects feedback errors. The proposed dual-timescale approach yields a negligible loss of spectral efficiency compared to a genie-aided scheme operating under error-free feedback and foreknown mobility model. Finally, a low-complexity policy is proposed by reducing the POMDP to an error-robust MDP. It is shown that the PBVI-and error-robust MDP-based policies improve the spectral efficiency by 85% and 67%, respectively, over a policy that scans exhaustively over the dominant beam pairs, and by 16% and 7%, respectively, over a state-of-the-art POMDP policy. I. INTRODUCTIONMillimeter-wave (mm-wave) has emerged as the most promising technology to enable multigigabit communication in vehicular communications, thanks to large bandwidth availability [2].By operating at carrier frequencies ranging from 30 GHz up to hundreds of GHz, mm-wave communications suffer from high isotropic path loss, overcome by using large antenna arrays

show abstract

Section: Simulation Setupmentioning

confidence: 99%

Learning and Adaptation for Millimeter-Wave Beam Tracking and Training: a Dual Timescale Variational Framework

Hussain¹,

Michelusi²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…transmitted by the UE to the BS in the last slot of the DT phase (at time t = k + T DT − 1). As in (11) for the BT feedback, Y = j denotes active SBPI detected, whereas Y = ∅ denotes inactive SBPI detection, due to either loss of alignment or blockage. Similarly to (11), = B I at the 2nd last slot (t = k + T DT −2) can be computed as a special case of ( 14) and ( 15) with |S BT | = 1 (since in the DT phase only one beam j is used for data transmission) and κL in place of L (since only κL symbols are used as pilot signal), yielding the probability of incorrectly detecting an active SBPI as…”

Section: E Beam-training (Bt) and Data Transmission (Dt)mentioning

confidence: 99%

“…In contrast, we contend and demonstrate numerically that learning and exploiting such information via adaptive communications can greatly improve the performance of mm-wave networks [10]. In our previous work [6], we bridged this gap by leveraging worst-case mobility information to design beam-sweeping and data communication schemes; in [11], we designed adaptive strategies for BT/DT that leverage a Markovian mobility model via POMDPs, but with no consideration of blockage (hence no handover).…”

mentioning

confidence: 96%

Mobility and Blockage-Aware Communications in Millimeter-Wave Vehicular Networks

Hussain

Scalabrin

Rossi

et al. 2020

IEEE Trans. Veh. Technol.

Self Cite

View full text Add to dashboard Cite

Mobility may degrade the performance of nextgeneration vehicular networks operating at the millimeter-wave spectrum: frequent mis-alignment and blockages require repeated beam-training and handover, with enormous overhead. Nevertheless, mobility induces temporal correlations in the communication beams and in blockage events. In this paper, an adaptive design is proposed, that learns and exploits these temporal correlations to reduce the beam-training overhead and make handover decisions. At each time-slot, the serving base station (BS) decides to perform either beam-training, data communication, or handover, under uncertainty in the system state. The decision problem is cast as a partially observable Markov decision process, with the goal to maximize the throughput delivered to the user, under an average power constraint. To address the high-dimensional optimization, an approximate constrained point-based value iteration (C-PBVI) method is developed, which simultaneously optimizes the primal and dual functions to meet the power constraint. Numerical results demonstrate a good match between the analysis and a simulation based on 2D mobility and 3D analog beamforming via uniform planar arrays at both BSs and UE, and reveal that C-PBVI performs near-optimally, and outperforms a baseline scheme with periodic beam-training by 38% in spectral efficiency. Motivated by the structure of C-PBVI, two heuristics are proposed, that trade complexity with sub-optimality, and achieve only 4% and 15% loss in spectral efficiency. Finally, the effect of mobility and multiple users on blockage dynamics is evaluated numerically, demonstrating superior performance over the baseline scheme. Index Terms-Blockage, beam training, handover, millimeter wave, Markov decision process, point based value iteration. I. INTRODUCTION C URRENT sub-6 GHz vehicular communication systems cannot support the demand of future applications such as autonomous driving, augmented reality and infotainment, due to limited bandwidth availability [2]. To this end, new Manuscript

show abstract

“…Sq ℓ,k ← Sq ℓ,k + γ 2 20: [6][7][8][9][10][11][12][13], which is the best optimistic estimate of the average rewards of the nodes. The second part is the beam measurement for the selected node (line 14).…”

mentioning

confidence: 99%

Online Learning for Position-Aided Millimeter Wave Beam Training

et al. 2019

View full text Add to dashboard Cite

Accurate beam alignment is essential for beambased millimeter wave communications. Conventional beam sweeping solutions often have large overhead, which is unacceptable for mobile applications like vehicle-to-everything. Learningbased solutions that leverage sensor data like position to identify good beam directions are one approach to reduce the overhead. Most existing solutions, though, are supervised-learning where the training data is collected beforehand. In this paper, we use a multi-armed bandit framework to develop online learning algorithms for beam pair selection and refinement. The beam pair selection algorithm learns coarse beam directions in some predefined beam codebook, e.g., in discrete angles separated by the 3dB beamwidths. The beam refinement fine-tunes the identified directions to match the peak of the power angular spectrum at that position. The beam pair selection uses the upper confidence bound (UCB) with a newly proposed riskaware feature, while the beam refinement uses a modified optimistic optimization algorithm. The proposed algorithms learn to recommend good beam pairs quickly. When using 16x16 arrays at both the transmitter and receiver, it can achieve on average 1dB gain over the exhaustive search (over 271x271 beam pairs) on the unrefined codebook within 100 time-steps with a training budget of only 30 beam pairs.

show abstract

Beam Training and Data Transmission Optimization in Millimeter-Wave Vehicular Networks

Cited by 16 publications

References 15 publications

Learning and Adaptation for Millimeter-Wave Beam Tracking and Training: a Dual Timescale Variational Framework

Learning and Adaptation for Millimeter-Wave Beam Tracking and Training: a Dual Timescale Variational Framework

Mobility and Blockage-Aware Communications in Millimeter-Wave Vehicular Networks

Online Learning for Position-Aided Millimeter Wave Beam Training

Contact Info

Product

Resources

About