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

Thanks to its capability of classifying complex phenomena without explicitmodeling, deep learning (DL) has been demonstrated to be a key enabler ofWireless Signal Classification (WSC). Although DL can achieve a very highaccuracy under certain conditions, recent research has unveiled that thewireless channel can disrupt the features learned by the DL model duringtraining, thus drastically reducing the classification performance inreal-world live settings. Since retraining classifiers is cumbersome afterdeployment, existing work has leveraged the usage of carefully-tailored FiniteImpulse Response (FIR) filters that, when applied at the transmitter's side,can restore the features that are lost because of the the channel actions,i.e., waveform synthesis. However, these approaches compute FIRs using offlineoptimization strategies, which limits their efficacy in highly-dynamic channelsettings. In this paper, we improve the state of the art by proposing Chares, aDeep Reinforcement Learning (DRL)-based framework for channel-resilientadaptive waveform synthesis. Chares adapts to new and unseen channel conditionsby optimally computing through DRL the FIRs in real-time. Chares is a DRL agentwhose architecture is-based upon the Twin Delayed Deep Deterministic PolicyGradients (TD3), which requires minimal feedback from the receiver and exploresa continuous action space. Chares has been extensively evaluated on twowell-known datasets. We have also evaluated the real-time latency of Chareswith an implementation on field-programmable gate array (FPGA). Results showthat Chares increases the accuracy up to 4.1x when no waveform synthesis isperformed, by 1.9x with respect to existing work, and can compute new actionswithin 41us.

Can You Fix My Neural Network? Real-Time Adaptive Waveform Synthesis for Resilient Wireless Signal Classification