Sept. 2023: Welcome to our new PhD student Changgyu Lee who joined us from Kyung Hee University, South Korea.
Sept. 2023: Welcome to our new PhD student Reza Sayyari who joined us from the University of Tabriz, Iran.
Feb. 2023: Byungju has accepted the position as an Assistant Professor at Pukyong National University, Busan, Korea. He will start in March. Congratulations Professor Lim!
Jan. 2023: Qing Lyu arrived at our lab as a PhD student. Welcome Qing!
Dec. 2022: Seok-Hyun completed his visit with us and will return to finish his PhD degree at Korea University.
Nov 2022: Prof. Vu joined the IEEE Transactions on Communications Editorial Board as an Editor.
Oct 2022: Prof. Young-Chai Ko visited LiNKS lab and gave a talk on inter-satellite communications.
Sept 2022: Beier Li officially started her PhD study, although she is no stranger to our lab as she just completed her MS degree at Tufts. Welcome Beier!
May 2022: Visiting PhD student Seok-Hyun Yoon is joining us from Prof. Young-Chai Ko’s research group in Korea University. Welcome Seok-Hyun.
April 2022: Dr. Byungju Lim is our postdoctoral scholar joining LiNKS from Korea University. Welcome Dr. Lim!
Spring 2021: Prof. Vu organizes a Tripods Spring seminar series on machine learning and data science in wireless communications, with potential applications in 5G and beyond-5G systems, IoT networks, and personal wireless devices.
Recent Publications
Tackling user association and handover in a dense 5G/6G network, we proposed the use of multi-agent Q-learning to adapt and perform the action in real-time. We designed two multi-agent policies, one centralized and one fully distributed, to handle the challenging issue of load balancing among all base stations and/or access points, as load balancing introduces dependency among different users’ actions. Our proposed algorithms show excellent adaptation to user mobility by achieving a higher network throughput at significantly lower handover rates than current 5G standards. The results are reported in the paper “Multi-Agent Q-Learning for Real-Time Load Balancing User Association and Handover in Mobile Networks,” recently accepted by the IEEE Transactions on Wireless Communications.
With an eye towards 6G, we analyzed conditions for coexistence between terrestrial networks and passive satellite services and derived bounds and tradeoffs between terrestrial node density and node transmitted power to achieve near-zero outage at a passive satellite receiver. We also derived conditions for the filters in terrestrial transmitters for satisfying out-of-band interference to the satellite. These results are reported in the paper “Interference Analysis for Coexistence of Terrestrial Networks with Satellite Services” published in the IEEE Transactions on Wireless Communications.
Our collaborative paper “Joint User Selection and Beamforming Design for Multi-IRS aided Internet-of-Things Networks” has been accepted to the IEEE Transactions on Vehicular Technology. The paper offers high performance and low complexity solutions to the design of active and passive beamforming jointly with the selection of high throughput users in a multi-IRS environment.
We showed that a graph neural network structure can be efficiently designed and trained in an unsupervised manner to effectively adapt base station beamforming and control the phase shift of multiple RISs (reconfigurable intelligent surfaces) in order to achieve high data rate that scales seamlessly with the network size and configuration. Our initial results will be published in “Graph Neural Network Based Beamforming and RIS Reflection Design in A Multi-RIS Assisted Wireless Network” at the 23rd IEEE Statistical Signal Processing (SSP).
We designed a novel and fully distributed multi-agent deep reinforcement learning (MADRL) algorithm using a deep Q-network structure and a multi-agent matching policy to ensure base station load balancing in our recent paper “Low Complexity Joint User Association, Beamforming and RIS Reflection Optimization for Load Balancing in a Multi-RIS Assisted Network”, published in the IEEE Wireless Communications Letters.
Our paper “Reinforcement Learning for User Association and Handover in mmWave-enabled Networks” has been accepted for publication in the IEEE Transactions on Wireless Communications. In the paper we designed new, efficient multi-arm bandit reinforcement learning algorithms for association and handover in a mobile wireless network, achieving higher network transmission sum rate than conventional method at a fraction of the handover rate, all the while maintaining the load balancing at each BS.
Our work on Energy-efficient Joint Wireless Charging and Computation Offloading In MEC Systems to minimize the total energy consumption while performing computation and charging the largest feasible amount of energy to the wireless end user will appear in an upcoming special issue in the IEEE Journal of Selected Topics in Signal Processing.
