Projects

2025

AgriLink: Internet of Plants for Smart Agriculture

Dense vegetation severely attenuates 2.4–5 GHz RF links (additional foliage loss of order 1–3 dB/m, and often >30 dB across canopies), forcing higher transmit power or more relays in plant/greenhouse settings. In contrast, Plant-Body Communication (PBC) utilizes signals within the stem, where ionic pathways and distributed capacitances form a guided, low-radiation medium, thereby reducing path loss and environmental variability. We target two practical outcomes: 1. quantify and model the intra-plant channel across coupling modes (galvanic vs. capacitive), species (herbaceous Dieffenbachia vs
Overview of wearable sensing technologies for comprehensive human health monitoring.

BioLink: Internet of Body via HBC for Healthcare

Human Body Communication (HBC) is an innovative technology that uses the human body itself as a transmission medium for data exchange. This approach enables secure, energy-efficient, and reliable communication between smart health sensors, wearable devices, and medical implants in and around the body. Unlike traditional wireless methods that rely on radio frequency signals prone to interference and privacy risks, HBC offers low power consumption, reduced signal loss, and enhanced data privacy by transmitting signals through direct skin or body tissue contact. This project focuses on advancing
digital-twin-for-indoor-scenario

Digital Twin For Indoor Scenario

In indoor wireless scenarios, Digital Twin enables the accurate modeling of complex environments, such as offices, factories, and smart homes, allowing for the real-time optimization of coverage, interference management, and seamless connectivity for IoT devices. IT supports applications like intelligent building management, immersive AR/VR, and industrial automation. At CCSL, we investigate two specific indoor DT scenarios: digital twin–aided beamforming design and digital twin–aided blockage prediction/detection for MIMO systems. Experiment Description Digital Twin (DT) technology enables
digital-twin-for-outdoor-scenario

Digital Twin For Outdoor Scenario

In outdoor scenarios, DTs provide large-scale virtual replicas of urban or rural networks to optimize base station placement, manage mobility, and enhance vehicular or drone communications. These advantages make DTs a powerful tool for improving reliability, efficiency, and adaptability across diverse wireless environments. At CCSL, we investigate two specific DT scenarios for outdoor wireless cellular networks: digital twin–aided beamforming design and digital twin–aided blockage prediction/detection for MIMO systems. Experiment Description Digital Twin (DT) technology enables the accurate

Digital Twin For RIS-Aided Systems

Beyond 5G networks are expected to deliver ultra-reliable, low-latency, and high-capacity connectivity. One of the key enabling technologies for achieving these goals is a reconfigurable intelligent surface (RIS), which is a programmable metasurface composed of many passive elements, each capable of adjusting the phase of incident signals. By intelligently controlling these elements, RIS can manipulate wireless propagation environment to enhance coverage, increase capacity, and improve energy efficiency. To accurately model the behavior of RIS within complex and dynamic environments, digital

EdgeCIM: A Hardware-Software Co-Design for CIM-Based Acceleration of Small Language Models

The deployment of language models is rapidly shifting from datacenters to edge devices such as laptops, smartphones, and embedded platforms, driven by the demand for interactive, low-latency, and privacy-preserving applications. In this context, Small Language Models (SLMs) have emerged as practical candidates, yet their inference reveals inefficiencies in conventional accelerators. While GPUs and NPUs process the GEMM-heavy prefill stage efficiently, they remain underutilized during the GEMV-dominated decoding phase, resulting in limited throughput and excessive energy consumption at the edge

Efficient AI Across Edge, Near-Edge, and Cloud

Modern applications like smart cameras, self-driving cars, and VR devices rely on powerful AI models. Running these models quickly and efficiently across phones, edge devices, and cloud servers is a tough challenge. Our work develops two frameworks to make this possible: DONNA finds the best way to split and run AI models across different types of devices, from traditional CPUs and GPUs to new Compute-In-Memory (CIM) accelerators, so they use less energy while staying fast. HiDist takes the idea further by looking at the whole system: edge devices near the user, stronger near-edge servers, and
ccsl-multi-target-detection-and-JVAR-estimation

Multi-Target Detection and JVAR Estimation

Integrated sensing and communication has been a key element towards the advancements of next generation wireless communication systems. Moreover, the convergence of radar and communication systems is rapidly gaining momentum as we advance towards beyond fifth-generation (5G) networks. A promising approach involves the utilization of existing 5G new radio 5G-NR control signals for sensing purposes. This approach exploits the periodic nature of such signals to collect environmental data without degrading the communication performance. Synchronization signal block (SSB) is the most popular
Testbed

Multimodal Sensing for Indoor RIS-Aided Beam Selection

In the 5G standard, beam sweeping is typically periodically executed through exhaustive search methods to maintain continuous alignment of user equipment (UE) and base station (BS) beamformers. In a reconfigurable intelligent surface (RIS)-aided massive MIMO (mMIMO) system, beam training is a daunting challenge as RIS codebook sizes are significantly larger compared to regular transceivers. To address the beam-training overhead, deep learning-based solutions have attracted great interest, enabling learning from data and adapting to dynamic conditions. The integration of multimodal sensing
Wearable health monitoring system with edge analytics for real-time diagnostics, emergency alerts, and secure cloud reporting.

On-Body HBC Networking

To advance innovations in next-generation Internet of Bodies (IoB) networks, there is a growing need for miniature, ultra-low-power devices capable of continuous and reliable operation. This need has led to a surge of interest in Human Body Communication (HBC), where nodes exchange messages directly through the human body as a secure and energy-efficient medium. With evidence showing that HBC can achieve up to 100× lower power consumption compared to conventional RF technologies, our research focuses on developing novel communication and networking techniques that facilitate connectivity over

Polar Codebook Design for Near-Field Communication

As sixth-generation (6G) wireless networks approach, leveraging the mmWave and THz bands' abundant spectrum becomes crucial, promising ultra-high data rates and enabling immersive communication experiences. This transformation, characterized by the integration of ultra-massive MIMO (UM-MIMO) systems, facilitates significant increases in throughput and capacity by utilizing densely packed antenna arrays. The resulting shift from traditional far-field communication, with its planar wavefronts, to near-field communication, where spherical wavefronts predominate, necessitates a reevaluation of
System overview of the RIS-aided indoor localization framework using RSSI and deep learning for precise positioning in NLoS environments.

RIS-Aided mmWave Indoor Localization

Deep Learning Framework for RSSI-Based Indoor Localization in RIS-Aided mmWave Systems. We present a novel deep learning framework for indoor localization in millimeter-wave (mmWave) environments using received signal strength indicator (RSSI) measurements from a reconfigurable intelligent surface (RIS)-aided system. Our dual-stream orientation-gated (DSOG) architecture addresses the critical challenges of non-line-of-sight (NLoS) conditions and orientation variability, achieving unprecedented decimeter-level accuracy with a median error of 0.19 m—outperforming classical methods by 74.7% and
Close-up view of a microfabricated sensor chip used in advanced biomedical and communication systems.

Ultra-Low-Power HBC Transceiver ICs

The CCSL advances next-generation Human Body Communication (HBC) through innovative ASIC implementations that enable ultra-low power, secure, and efficient data transmission using the human body as a communication medium. Our research focuses on silicon-proven solutions that bridge the gap between theoretical HBC concepts and practical wearable/implantable applications. We develop mixed-signal transceivers that support multiple communication modes, adaptive signal processing, and AI-enhanced performance optimization. The lab's ASIC implementations demonstrate >100x energy efficiency