Shannon Wireless’ Post

#RecommendedPaper Fault Diagnosis in Analog Circuits Using Swarm Intelligence by Nadia Nedjah, et al. ➡️Read the full paper here: https://lnkd.in/dnKnqfTJ #analog #circuit #SwarmIntelligence #optimization #algorithm

Practice Question on "Digital Signal Processing" Discrete-time Fourier transform computation. #Fourier #Discrete_time #Digital_Signal_Processing

New Post: Optical Flow: Lucas-Kanade Method

While doing CTPR&D (https://lnkd.in/ejzqr_HA) the other day, I was thinking about optical tech for future #CTPcraft and how to 'tie-in' optical and digital tech data via CTP to C-domain for FTL C-signal propagation of data along the C-energy. I was already thinking along these lines this article is citing. Using 'light' P-energy makes more sense in a way for even computer applications, as light-photons going through a photonic-medium would likely have less Ohms resistance than raw electron current traveling over a copper circuit over a PCB. And Voilà! Eureka! Will be interesting to see if we can ever utilize light for CTP-related applications. Interesting CTPSci FACT: while light photons have no mass, physical light is still a form of the P-energy. (https://lnkd.in/eeQZrwUK) #newscience #processors #optics #opticalprocessors

This paper leverages deep semisupervised learning for #radio #frequency (#RF) #fingerprinting, which largely relies on a composite data augmentation scheme specifically designed for wireless communication signals, combined with two popular techniques: 1) #consistency-#based #regularization and 2) #pseudo-#labeling. Experimental results on both simulated and real-world datasets demonstrate that our proposed method for semisupervised RF fingerprinting is far superior to other competing ones, and it achieves remarkable performance almost close to that of fully supervised learning, with a very limited number of examples available. ----Weidong Wang, Cheng Luo, Jiancheng An, Lu Gan More details can be found at this link: https://lnkd.in/g89tJyhP

This Voltus InsightAI blog examines the intricate technologies that make AI-driven EM-IR analysis and closure possible. https://ow.ly/h2gh30sCZOr

In this letter, the authors propose two unsupervised #deep #neural #networks (#DNN) architectures, fully and partially distributed, that can perform decentralized coordinated beamforming with zero or limited communication overhead between #access #points (#APs) and #network #controller (#NC), for both fully digital and hybrid precoding. The proposed DNNs achieve near-optimal sum-rate while also reducing complexity by 10-24\times compared to conventional near-optimal solutions. ---- Hamed Hojatian, Ph.D., Jeremy Nadal, Jean-Francois Frigon, François Leduc-Primeau More details can be found at this link: https://lnkd.in/eAaSc7_m

Happy to announce that my paper "Automated Generation of Transformations to Mitigate Sensor Hardware Migration in ADS" was accepted to the high-impact journal IEEE Robotics and Automation Letters and is available as a preprint via IEEE Early Access! Autonomous driving systems (ADSs) rely on massive amounts of sensed data to train their underlying deep neural networks (DNNs). Common sensor hardware migrations can render an existing DNN-dependent pipeline inadequate. This necessitates the development of bespoke transformations to adapt new sensor data to the old trained network, or the retraining of a new network with new sensor data. These solutions are expensive, often performed reactively to sensor hardware migration, and rely only on empirical reconstruction and validation metrics which lack knowledge of the features important to the trained DNN. To address these challenges, we propose PreFixer, a technique that can systematically generate transformations for many types of sensor hardware migration during the ADS development lifecycle. PreFixer collects small datasets using colocated new and old sensors, and then uses that data and the output of the original trained DNN to train an augmented encoder to learn a transformation that maps new sensor data to old sensor data. The trained encoder can then be deployed as a preprocessor to the original trained DNN. Our study shows that, for a common set of camera sensor hardware migrations, PreFixer can match or improve the performance of the best-performing specialized baseline technique in terms of distance travelled safely with 10% of the training dataset, and take at most half of the training time of a new network. Many thanks to my co-author and advisor Sebastian Elbaum and co-author Hongning Wang! And many thanks to the reviewers and editor of RA-L Ashis Banerjee for your support in improving and publishing this paper. DOI: 10.1109/LRA.2024.3405810 RA-L Preprint: https://lnkd.in/g_VXasxF

No instrument can measure reflectivity and transmittance at the same time except InspectRx® reflectivity, SpectRx™ absorption rate, computing transmittance rate = 1 - reflectivity - absorption rate, and algorithm. https://lnkd.in/guXXQFEq

LIVE show of Elektor Lab Talk coming up today at 4 PM (CEST) Tune in https://ow.ly/pqyq50PwBR9 Join Elektor Editors as they talk about AI and the new Circuit Special of Elektor Magazine! #elektor #electronics #ai