News
FuNCtions lab is part of a $24M ARO ICE-CEM Center led by Duke focusing on inducing cyber and physical effects in unmanned aerial systems via cyber-EM convergence.
Prof. Chen is a co-PI of a 3-year, $4.1M NSF VINES Track 2 award on AReS, which aims to assess, restore, and support post-disaster connectivity and situational awareness through AI-enabled ISAC and ISAC-AR integration. This project is led by Princeton University, in collaboration with UC Santa Barbara, UW-Madison, and Ericsson.
A paper on in-situ radar sensing on 5G base stations with zero-shot template generation was accepted to ACM MobiHoc’26. [arXiv]
FuNCtions lab received a grant from ARO to work on real-time, scalable, and programmable radio architectures based on the RFSoC platforms leveraging AI and agentic workflows.
The NSF AI Institute for Edge Computing and Systems (Athena), led by Duke, was renewed with a 5-year, $15M award to pursue “Immersive AI in the Physical World” across computing systems, foundational AI, physical AI, and applications.
Prof. Chen is the PI of a $9M grant from ARO focused on RAPTORS–Rapid Analog Processing for Tactical Observation and Opportunistic Radio Spectrum Access. In collaboration with Professors Miroslav Pajic and David Smith (Duke), Arun Natarajan (Yale), Aravind Nagulu (Northeastern), and Raytheon BBN Technologies, this project aims to build intelligent sensing radios and networks with low SWaP-C.
A paper on optimizing arbitrary RF containment using tactical aerial networks and differentiable ray tracing was accepted to IEEE MILCOM’26.
Prof. Chen is the PI of an NSF VINES Track 1 Award focused on building photonics-enabled, energy-efficient massive MIMO radios. This project is in collaboration with Professors Ali Niknejad, Jun-Chau Chien, and Zaijun Chen from UC Berkeley.
A paper on 5G mmWave ISAC through sub-symbol beam switching with CSI extraction and reconstruction was accepted to IEEE Transactions on Mobile Computing (TMC). A preliminary version of the work was also presented at IEEE MTT-S RFSA’26.
One paper on fiber sensing-based borehole seismic velocity profiling was accepted to ECOC’26. This paper was based on a field trial we conducted at the Newberry Well, OR, and in collaboration with Enegis LLC and NEC Labs America.
Several FuNCtions lab students are out for summer internships: Wei Cheng (Seagate), Mingzhe Han (NEC Labs America), Zhenzhou Qi (Samsung Research America), Sam Rivera (The Aerospace Corporation), and Zehao Wang (Keysight Technologies).
FuNCtions lab received a Thomas Lord Educational Innovation Grant Program award from Duke’s Pratt School of Engineering, which will support the expansion of my full-stack IoT systems course to incorporate emerging edge AI and physical AI topics and hands-on labs.
Zhihui Gao (co-advised with Prof. Yiran Chen) successfully defended his Ph.D. dissertation and will join MIT RLE as a postdoc. Congratulations, Dr. Gao!
Yiming Li received the ECE Outstanding Graduate Teaching Assistant Awards for the 2025–2026 academic year! Yiming served as the teaching assistant in my ECE 655/CompSci 655 “Full-Stack IoT Systems” course. Congratulations Yiming!
Prof. Chen was invited to attend the AI-RAN 2026 May Member Meeting in Chandler, AZ, and participated in panel on “Data-for-AI (D-4-AI) Academic Perspectives”.
Prof. Chen organized the 1st Advanced Wireless and Edge AI Workshop in the Research Triangle, co-hosted by Duke Engineering and Keysight Technologies, bringing together faculty, industry leaders, students, and regional collaborators for a day of discussion around the future of wireless systems and edge AI. [Duke ECE news]
Sam Rivera received the pretigious National Defense Science and Engineering Graduate (NDSEG) fellowship, which supports outstanding graduate students pursuing cutting-edge research in science and engineering. Congratulations Sam!
One paper on skilled AI agents for embedded and IoT systems development was accepted to the System Demonstrations Track at ACM CAIS’26. [tech report] [Code and Benchmark]
Three papers papers were accepted to IEEE DySPAN’26: (i) Multi-channel real-time D-band beamformer based on the RFSoC ZCU216 platform, (ii) Real-time imaging processing for spectrum energy detction, and (iii) Radio dynamic zone management system design for the PAWR COSMOS testbed.
Our paper on simultaneous radar-based odometry and synthetic-array sensing for UAVs was accepted to IEEE ICRA’26.
Our paper on mmWave ISAC with sub-symbol beam switching was accepted to IEEE MTT-S RFSA’26.
Our paper in collaboration with the Duke Quantum Center on the first coexistence field Trial of 1092 nm quantum link, coherent 400 GbE, and 5G Service was accepted to IEEE/Optica OFC’26.
Our education mateiral on the design and control of IoT systems via LLMs using MCP was selected as a contributed talk at the Education Track at NeruIPS 2025. Please also check our our poster and booth at this year’s NeurIPS! [techical paper at ACM WiNTECH’25]
Our paper on disaggregated deep learning using in-physics computing directly at RF was accepted to Science Advances. In this paper, we introduced and demonstrated a wireless edge inference architecture that enables disaggregates model access by broadcasting ML model weights over the air and performs complex-valued matrix-vector multiplications directly at the edge devices in the RF domain. This series of work is in collboration with Prof. Dirk Englund’s group at MIT. We will also present this work at the Workshop on AI and ML for Next-Generation Wireless Communications and Networking (AI4NextG) at NeruIPS 2025. [experimental paper] [theoretical paper]
Our paper on efficient and scalable multi-cell mmWave baseband processing with heterogeneous compute (CPUs and ASICs) was conditionally accepted to ACM MobiCom’26. [tech report]
Prof. Chen was invited to attend the Brooklyn 6G Summit and honored to moderated a panel on “Deep dive: Value creation with AI” with leaders from both industry and academia: Kaushik Chowdhury (UT Austin), Harish Viswanathan (Nokia Bell Labs), Mikael Hook (Ericsson Research), Udayan Mukherjee (Intel), and Doru Calin (MediaTek).
Together with Dan Kilper and Camille Delezoide, we are excited to launch the Inaugural Dataset Paper Submission at OFC 2026, which aims to solicit dataset papers and to establish a collaborative platform for researchers, scholars, and practitioners from academia and industry to develop, evaluate, and showcase ML techniques applied to diverse datasets collected from real-world optical devices, platforms, and networks. Please consider submitting your dataset papers!
FuNCtions lab is part of an NSF CIRC GRAND/ENS Award (PI: Ivan Seskar at Rutgers University) to expand and enhance the PAWR COSMOS testbed deployed in Upper West Manhattan, New York City, enabling research in the areas of 6G wireless networks, smart cities, edge AI, and optical networking.
Two papers papers were accepted to IEEE MILCOM’25: Multi-channel real-time SDR based on the RFSoC ZCU216 platform in collaboration with Prof. Ali Niknejad’s group, and over-the-air experimentation with Zak-OTFS waveforms using spread pilot in collaboration with Prof. Robert Calderbank’s group.
Prof. Chen is the PI of a grant from ARO (Co-PI: Prof. Dirk Englund from MIT) to work on enabling machine intelligence on wireless networks (MIWEN) via ultra-low-power in-physics computing using RF signals and electronics. This is based on our recent work on enabling in-physics computing for deep learning directly at radio frequency, with proof-of-concept results demonstrated in this paper.
FuNCtions lab received a research grant from NEC Labs America to work on open optical transport system for data center appliations. This project will leverage the programmable Duke BlueFrog optical testbed deployed across the Research Triangle.
Prof. Chen received the Nortel Networks Professorship from Duke Engineering.
FuNCtions lab is part of an NSF CC* Integration-Large Award (PI: Prof. Danyang Zhuo, in collaboration with Duke OIT) to design and deploy an RDMA-based campus network that enables the shared use of distributed, heterogeneous GPUs to accelerate scientific applications and fast access to research data storage.
Prof. Chen received the NSF CAREER Award to work on enhancing the energy efficiency and scalability of mobile fronthaul networks using analog radio-over-fiber (ARoF) technologies. This project will also leverage the programmable wireless-optical testbeds, including the Duke BlueFrog and PAWR COSMOS platforms, to explore new networks capabilties supported by ARoF.
An extension of our ACM MobiHoc 2023 paper on the optimization, scheduling, and stability analysis of sectorized wireless networks was accepted to IEEE/ACM Transactions on Networking (TON).
A paper on our developed light-weight method for enhancing radar point cloud using graph neural networks, particularly in complex and dynamic environments, was accepted to IEEE/RSJ IROS’25.
Wei Cheng and Zehao Wang received the ECE Outstanding Graduate Teaching Assistant Awards for the 2024–2025 academic year! Wei and Zehao served as the teaching assistants in my ECE 655/CompSci 655 “Full-Stack IoT Systems” course. Congratulations Wei and Zehao!
Prof. Francesco Gringoli and I are co-charing ACM WiNTECH 2025, co-located with ACM MobiCom 2025. Please consider submitting your work!
Functions lab received a research grant from NTT to work on machine learnign tehcnology for resilient optical networks.
FuNCtions lab received an NVIDIA Academic Grant to work on joint communication and sensing leveraging high-fidelity wireless digital twins built using our recently developed Geo2SigMap framework.
A paper on our developed digital twin framework and generalized deep learning model for signal coverage prediction was accepted to IEEE International Symposium on Dynamic Spectrum Access Networks (DySPAN).
A paper on our experiences with building full-duplex radios and their integration in the PAWR COSMOS testbed was accepted to IEEE Transactions on Wireless Communications (TWC).
Two papers were accepted to OFC’25: (i) Multi-span OSNR and GSNR prediction using cascaded learning (oral), and (ii) Scalable ML models for optical transmission system management (invited talk).
FuNCtions lab received an NVIDIA Academic Grant to work on building high-fidelity 3D world models and wireless digital twins leveraging our recently developed Geo2SigMap framework.
FuNCtions Lab is at the ACM MobiCom’24 in Washington, D.C., where we presented two full papers, two workshop papers, five demos, and one poster. [blog post]
FuNCtions lab received a Duke Science and Technology (DST) Launch Seed grant to work on building the Duke quantum networks with colleagues from the Duke Quantum Center.
Our paper on real-time, efficient baseband processing for mmWave networks was accepted to ACM MobiCom’24. [code]
Prof. Chen is the PI of an award from the NSF NewSpectrum program to work on dense and in-situ spectrum monitoring leveraging efficient circuit-system co-design. This project is collaborative project with Professors Monisha Ghosh (U. Notre Dame), Arun Natarajan (Yale), and Aravind Nagulu (Northeastern).
Prof. Chen gave an invited talk at the OPTICA webinar about our work on optical network performance modeling and optimization. [link to talk video]
FuNCtions lab received a grant from Duke Engineering’s Beyond the Horizon Initiative to develop a dark fiber network for hybrid quantum-classical experimental research. This is a collaborative project with the Duke Quantum Center (DQC) and Duke Office of Information Technology (OIT).
Zeyu (Michael) Li received the 3rd Place Winner in the Spring 2024 ECE Independent Study Poster Session.
One postdeadline paper accepted to OFC’24 on the first field demonstration of 4D optical link tomography using transponder capable of distance, time, frequency, and polarization-resolved monitoring.
Our OFC’24 paper on multi-span optical power spectrum prediction using cascaded learning was selected by the subcommittee as a Top-Scored Paper. Zehao was also selected as a finalist for the Corning Outstanding Student Paper Competition.
Our paper on optimized analog multi-user beamforming in mmWave networks was accepted to ACM MobiCom’24.
Our paper on RF signal mapping using public geographic databases (OpenStreetMap), open-source ray tracing tools (Sionna RT), and ML was accepted to IEEE DySPAN’24. [code and dataset]
Our paper on real-time generation of high-fidelity point clouds using low-cost mmWave radars was accepted to IEEE ICRA’24.
Three papers were accepted to OFC’24: (i) Multi-span optical power spectrum prediction using cascaded learning, (ii) Transceiver BER-OSNR modeling for QoT estimation, and (iii) Fiber type identification using in-service Brillouin optical time domain analysis.
Zeyu (Michael) Li was selected as a Pratt Research Fellow (Class of 2025).
Our paper on a blackbox Physical layer attack framework on mmWave automotive FMCW radars was accepted to ISOC NDSS’24. [arXiv]
Our paper on automated and on-demand optical wavelength path provisioning for data center exchange services was selected as a Highest Scoring and Best Paper at ECOC’23. The results were also presented by NTT at the Telecom Infra Project’s (TIP) Fyuz event in Spain. [NTT press release]
Our paper on the field trial of coexistence of real-time fiber sensing and coherent 400 GbE signals was accepted to IEEE/Optica Journal of Lightwave Technology, Special Issue on Top-Scored Papers from IEEE/Optica OFC’23.
Prof. Chen is the PI of an NSF EAGER award that aims to build an integrated fiber sensing and communication testbed leveraging Duke-Durham field fiber. This project is in collaboration with the Duke Office of Information Technology (OIT).
FuNCtions lab received an NSF supplement to conduct experimental millimeter-wave research on the NSF-funded PAWR COSMOS testbed.
Two papers were accepted to ACM MobiHoc’23: (i) Wi-Fi monitoring using low sampling rate radios, and (ii) Optimization of sectorized wireless networks.
Our paper on the design and deployment of programmable millimeter-wave radios in the PAWR COSMOS testbed was accepted to Computer Networks.
FuNCtions lab received an award from the Thomas Lord Educational Innovation Grant Program that will support the development of an undergraduate course on full-stack IoT systems.
Our paper on an open EDFA gain spectrum dataset collected using the PAWR COSMOS testbed and its applications in data-driven EDFA gain modeling was accepted to IEEE/Optica Journal of Optical Communications and Networking. [COSMOS EDFA dataset]
Three papers were accepted to ECOC’23: (i) Field trial of automatic WDM optical path provisioning, (ii) One-shot transfer learning for EDFA gain spectrum prediction, and (iii) ML-based Raman tilt prediction in ROADM transmission systems.
FuNCtions lab received a research grant from NTT to work on autonomous WDM optical link control frameworks.
Our OFC’23 paper on the field trial of coexistence of real-time fiber sensing and coherent 400 GbE signals was selected by the subcommittee as a Top-Scored Paper.
FuNCtions lab is part of a $35M SRC JUMP 2.0 Center for Ubiquitous Connectivity (CUbiC) with the goal of enabling seamless, robust, and scalable edge-to-cloud connectivity by developing and integrating advanced photonic, wired, and wireless connectivity solutions. CUbiC consists of a team of world-class researchers from Columbia (lead institution, director: Prof. Keren Bergman), UC Berkeley, UCSB, UIUC, MIT, UMich, Cornell, Duke, Princeton, Stanford, Oregon State University, UCSD, and USC. [SRC JUMP 2.0 program] [DARPA news] [Columbia Engineering news] [Duke Pratt news]
FuNCtions lab is part of an award from the NSF SII-NRDZ program to work on spectrum sharing field trials in an urban FCC innovation zone. This is a collaborative project between Columbia (lead institute, PI: Prof. Gil Zussman), Syracuse, Princeton, Duke, Rutgers, and CCNY. [NSF award]
Two papers were accepted to OFC’23: (i) Field trial of coexistence and simultaneous switching of real-time fiber sensing and coherent 400 GbE, and (ii) EDFA wavelength-dependent gain prediction using transfer learning.
Prof. Chen is the PI of an NSF NeTS Medium award to develop systems supporting the softwarization of mmWave RANs at the Edge. This project is collaborative project with Prof. Lin Zhong from Yale.
Our paper on ML-based optical power spectrum prediction in multi-span ROADM systems was accepted to ECOC’22 (oral presentation).