Research

Recent Interests

Recently, our group has worked on the following topics:

  • Green Cellular Networks: Graduate students Rajshekhar Bhat, Luo Shixin, Jia Chenlong, and Guo Zheng have all worked on various aspects of the general problem of making wide area networks more energy efficient i.e. minimizing the amount of energy needed to deliver each bit of information. The mathematical tools needed in this work include optimization theory, random networks, and percolation theory. I was also the founding chair of the special interest group on Green Cellular Networks, within the IEEE Comsoc Technical Committee on Green Communications and Computing.
  • Smart Grid: When cellular infrastructure is powered by the smart grid, with time-varying energy prices and locally generated renewable energy, opportunities exist for network operators to lower their energy costs through the careful design of battery charging schedules, and real-time decisions on when to sell excess energy back to the utility. Graduate student Johann Leithon worked very successfully on this problem, co-supervised by Dr. Sun Sumei of the Institute for Infocomm Research (I2R).
  • IoT/M2M Communications and Security: Recent work by PhD students Ayush Kumar and Nalam Abhishek, and post-docs Anshoo Tandon, Utku Tefek and Chenlong Jia are working/have worked on this topic, mostly under the auspices of the SingTel-NUS Corporate Lab. Within this lab, we are experimenting with hardware implementations of feasible PHY or MAC layer cyber-security attacks and methods for defending against these. Utku Tefek has studied the use of stochastic geometry in analyzing M2M random networks, while post-docs Chen Zhi and Han Xiao designed relay networks and other types of wireless topologies with a cross-layer perspective e.g. accounting for queueing delays and queue stability while at the same time satisfying rate and/or power requirements.

Past Work

In the past, my research has spanned the following areas:

  • Adaptive Signal Processing (1992-1996): My PhD work was on the use of adaptive IIR filters in acoustic echo cancellation, where the long system impulse response created problems with convergence when using a conventional adaptive FIR filter.
  • Multi-User Detection (1996-2002): In most of this period, I was a researcher at the Centre for Wireless Communications (one of I2R's predecessors) where a small group led by Lars Rasmussen and sponsored by Oki Inc. of Japan worked intensely on the problem of interference mitigation in CDMA (code division multi-access) networks. A number of highly cited papers resulted from this project, including the ones on multi-stage interference cancellation, and Kalman filter-based CDMA detection.
  • Multi-Antenna Systems (2001-2010): At U of T, we worked on several problems in the multi-antenna domain, including single- and multi-user transmitter precoding, and receiver design.
  • Relay Systems (2004-present): In a series of papers, PhD student Zhao Yi (co-supervised with Ravi Adve) made important contributions to the understanding of amplify-and-forward relay networks. These are still regularly cited today. With M.A.Sc. student Liu Xi, I also studied the design and performance of fountain codes in a relay system.
  • Variational Inference in Communications (2004-2008): PhD student Darryl Lin applied the variational inference technique to a number of communications problems, including multi-user detection and OFDM phase noise mitigation. This resulted in a number of very original and highly regarded papers.
  • Multi-Carrier Communications (2005-2010): When OFDM became the technology of choice in wireless communications, Toronto-based Redline Communications Inc. partnered with us to work on various aspects of OFDM receiver design, including phase noise mitigation and resource allocation. The resulting work by Darryl Lin, Taiwen Tang, Roya Doostnejad, Gokul Sridharan and others resulted in a number of publications in journals and conferences.
  • Cognitive Radio (2007-2011): We developed spectrum sensing algorithms for multi-antenna receivers based on the generalized likelihood ratio test (GLRT) in my sabbatical at I2R in 2008, the results of which were published in two highly cited papers. PhD student Sepideh Zarrin completed some useful work on cooperative sensing, and quickest detection-based sensing. Currently, we are interested in MAC protocols in cognitive radio for M2M applications.