The use of deep neural networks (DNNs) is currently transforming many areas of science and engineering. Although DNN-based techniques outperform traditional algorithms in most signal processing tasks, they can exhibit weaknesses such as reduced robustness and a tendency to produce hallucinations. These issues are linked to the DNN's Lipschitz constant, which typically worsens exponentially with the addition of layers. In this work, we present a framework for the design of stable networks with maximal expressivity.

The recent development of DNA-based data storage prototypes has raised several questions about how to optimally encode information in these systems. A distinguishing feature of this new storage paradigm is that the stored information is read via “shotgun” sequencing technologies. This means that the channel output comprises many short fragments of the input observed out of order. Motivated by this, we study the capacity of a class of “shuffling channels” that capture this inherent need to reorder the observed channel output.

This seminar provides key insights into wireless system analysis from a circuit designer’s perspective, focusing on modulation techniques, line-up analysis, and link performance optimization. Participants will explore impairments in analog and mixed-signal components-thermal noise, phase noise, spurious signals, and distortion and their impact on system performance. The session covers critical metrics such as Signal-to-Noise Ratio (SNR), Error Vector Magnitude (EVM), Receiver Sensitivity, Blocker performance, and Transmitter out-of-band noise.

High-resolution ADCs are essential components in many biomedical and environmental sensing applications. As the demands for wearable, miniature and point-of-care systems keep increasing, the design requirements for high-resolution ADCs also get more stringent, with strong emphasis on energy efficiency, reliability, and low cost. Over the past few years, many design techniques have been developed to excel the figure-of-merit (FoM) of high-resolution ADCs, where the state-of-the-arts has gotten really close to the theoretical limit.

ESPRIT-based high-resolution parameter estimation algorithms in DFT beamspace have been proposed as efficient gridless channel estimation schemes for MIMO OFDM systems in the millimeter wave (mmWave) band. Compared to conventional ESPRIT-based algorithms in element space, the DFT beamspace approach can be applied to MIMO systems with hybrid analog-digital architectures. Moreover, this approach significantly reduces the training overhead for communication systems operating in the mmWave band. It involves coarse and fine estimation steps.