Shiwen Mao

Excited to share two new IEEE Journal of Solid-State Circuits (JSSC) papers from our group, both pushing the boundaries of UWB wireless links for high-channel-count implantable neural interfaces! Over the past two years, we have been exploring how UWB can deliver high-data-rate, low-power, and robust wireless communication through biological tissue, and this effort has now resulted in two JSSC papers, each addressing a key challenge in next-generation neural implants: 1️⃣ A 3-mm² multi-data-rate IR-UWB TX with an on-chip antenna, achieving - Flexible, multi-rate operation: 50 Mb/s to 1.2 Gb/s - Ultra-compact 3-mm² footprint - Dual-modulation: 4-PPM + 2-PSK (high rate), OOK (low rate) - 1.49–7.26 mW power consumption across modes This is the first mm-scale UWB transmitter that combines Gb/s-class data rates, extended communication range, and full multi-rate flexibility, enabling high-density, multi-mode (e.g., closed-loop) neural interfaces. (extended from our ESSERC 2024 work) https://lnkd.in/ezdgjTfT 2️⃣ An IR-UWB TX with co-designed PA & antenna This paper introduces a 49.8-mm² IR-UWB transmitter with a co-designed power amplifier and antenna, achieving a unique combination of high data rate (500 Mb/s), extended transmission range (meter-scale), and miniaturized, mm-scale form factor. (extended from CICC 2024) https://lnkd.in/ean52xqq I am proud of Cong Ding for her outstanding work on these projects, and many thanks to our co-authors Mingxiang Gao and Prof. Anja Skrivervik at EPFL for the collaborative effort. We are also grateful for the support from the SwissChips initiative. #JSSC #BCI #Neurotechnology

512

8 Comments

🌟 A Strong Year of Growth for J-MMCT 🌟 Thanks to the collective efforts of our community, the journal achieved significant improvements across all key publication metrics in 2025: 📈 Submissions: +16% ⏱️ Submission → First Decision: −18% (9 days faster) ⏱️ Submission → Final Decision: −26% (25 days faster) ⏱️ Submission → Publication: −20% (30 days faster) 📥 Total Usage (Downloads): +11% 📚 Citations: +20% We will continue to publish innovative research in a timely manner that advances the field.

19

Quantum deep reinforcement learning (QDRL) for optimizing 6G satellite-air-ground integrated networks (SAGIN) with digital twin. Please check our newly accepted paper on IEEE IoT journal. This work is led by my PhD student Sasinda C. Prabhashana in collaboration with Berk Canberk, Haejoon Jung, Simon Cotton https://lnkd.in/eDqMnWnP

180

Long delayed post from http://icccn.org/ICCCN25/. Sepideh Abdollah recently published how to make multi-source downloads more efficient at IEEE ICCCN '25. In this work, she developed a Multi-Source Data Transfer Protocol (MDTP) - a method that enables file downloads from multiple servers in parallel. Thanks to Craig for all his valuable input. https://lnkd.in/eUsgdHwy The trick to better performance and more balanced transfers? Use all replicas (even the slow ones) and adapt request sizes to each server's observed throughput. MDTP splits a file into smaller chunks and distributes them across servers, adapting chunk sizes based on each server’s current performance. In our evaluation, MDTP reduced transfer times by 10–22% compared to Aria2, and showed even larger gains over static chunking and BitTorrent. Currently, Sepideh is extending this idea by integrating machine learning to make MDTP adaptive and predictive - allowing it to learn from previous transfers and adjust chunk sizes dynamically. Very excited to see what this can do for large-scale scientific data transfers in the near future.

29

6 Comments

Since IEEE/ACIS SNPD-2025 IV schedule is now out, I can now reveal more information about my paper. The paper is titled "Chatbots as Turing Machines" and it discusses machine learning models and chatbots as Turing Machines. The goal of the paper is to link modern LLM-based AI together with more classical computer science. I hope the paper will spark some discussions.

12

I'm happy to share our recent publication in IEEE Control Systems Letters, titled: Choosing Between Active and Passive Flow Control via Input-Output Analysis: Application to Couette Flow. This study provides a quick, physics-informed rule for designing actuation strategies without the need for costly experiments or DNS studies. We demonstrate our methodology by quantifying the optimal actuation frequency that yields the maximal response in plane Couette flow—a fundamental geometry in transition and flow control studies, which allows focus on purely shear-driven flows. We classify the conditions in which the frequency is zero—corresponding to passive control being preferable—and nonzero—corresponding to active control being preferable. Kudos to Ofek Frank-Shapir for his first journal publication! Link to the paper: https://lnkd.in/dGhxr8JR

56

6 Comments

I am pleased to share our latest publication titled "Stochastic Modeling of EV On-Board Chargers for Fast Frequency Response Under Communication Delays" in the IEEE Open Journal of Vehicular Technology. This work was led by my doctoral student Xiang Shi and was partially funded by the ETP Doctoral Program, with Arnold Clark as the industry partner. The paper proposes a stochastic model to evaluate the readiness of various EV models under communication delays, both hardware and over the air, for frequency services. Early access is available here: https://lnkd.in/ePte_p3A

45

1 Comment

🚀 Excited to share that our latest research has been published in the IEEE Control Systems Letters! 📄 "Stabilization of Nonlinear Systems via Subset-Based Membership Functions and Exact Polytopic Representation" 🔗https://lnkd.in/ga9vnJpx 💡In this work, we present a novel method to stabilize nonlinear systems using an exact polytopic representation within a region of interest. Unlike traditional methods, such as gain scheduling or piecewise approximations, the proposed approach does not rely on such approximations. Our design algorithm ensures asymptotic stability and ensures that the derived stabilization conditions are directly applicable to the original nonlinear system. 📊 Results: Numerical examples validate the robustness and accuracy of our method, showing real advantages over traditional approaches. 🤝 This research was developed in collaboration with Thierry Marie Guerra, Rafael Nascimento Silva, and Mohamed Diouf.

48

1 Comment

Our new paper was just online in IEEE Transactions Communications with titled "RIS-assisted SATINs with RSMA and DRL: A Trade-off between Spectral, Secrecy, and Energy Efficiency". In this paper, we address the problem of maximising secrecy energy efficiency (SEE) in SATINs, which can accurately describe the effective trade-off between security, spectral efficiency and transmit power. Particularly, we investigate a secure beamforming (BF) scheme in cognitive SATINs that employs rate-splitting multiple access (RSMA) and reconfigurable intelligent surface (RIS) in the presence of multiple eavesdroppers (Eves) in a UAV-aided secondary network (SN). To optimize the SEE for secondary vehicle users while satisfying the constraints of primary users (PUs), we utilize deep reinforcement learning (DRL) to address the coupling between different optimized parameters based on the improved long short-term memory proximal policy optimization (LSTM-PPO) algorithm. The main innovation of this paper is to design a sophisticated reward function, action space, and state space according to each constraint to speed up the convergence. In addition, simulation results show that the proposed DRL-based optimization scheme exhibits significant advantages in terms of SEE compared with benchmark schemes, validating the effectiveness of this work. https://lnkd.in/g3VizaHH

57

12 Comments