English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Millimeter wave (mmWave) and terahertz MIMO systems rely on pre-definedbeamforming codebooks for both initial access and data transmission. Beingpre-defined, however, these codebooks are commonly not optimized for specificenvironments, user distributions, and/or possible hardware impairments. Thisleads to large codebook sizes with high beam training overhead which increasesthe initial access/tracking latency and makes it hard for these systems tosupport highly mobile applications. To overcome these limitations, this paperdevelops a deep reinforcement learning framework that learns how to iterativelyoptimize the codebook beam patterns (shapes) relying only on the receive powermeasurements and without requiring any explicit channel knowledge. Thedeveloped model learns how to autonomously adapt the beam patterns to bestmatch the surrounding environment, user distribution, hardware impairments, andarray geometry. Further, this approach does not require any knowledge about thechannel, array geometry, RF hardware, or user positions. To reduce the learningtime, the proposed model designs a novel Wolpertinger-variant architecture thatis capable of efficiently searching for an optimal policy in a large discreteaction space, which is important for large antenna arrays with quantized phaseshifters. This complex-valued neural network architecture design respects thepractical RF hardware constraints such as the constant-modulus and quantizedphase shifter constraints. Simulation results based on the publicly availableDeepMIMO dataset confirm the ability of the developed framework to learnnear-optimal beam patterns for both line-of-sight (LOS) and non-LOS scenariosand for arrays with hardware impairments without requiring any channelknowledge.

Reinforcement Learning of Beam Codebooks in Millimeter Wave and Terahertz MIMO Systems