Research
Propagation channel measurement and modelling
Everything else depends on this. A statistical channel model is only as good as the measurement campaign behind it, and the environments worth modelling are usually the ones nobody has measured properly: confined spaces, heavy multipath, frequency bands where the equipment is new. Work here includes statistical modelling for the terahertz band, where path loss and absorption behave unlike anything at microwave frequencies.
Underground mine and disaster communications
A mine tunnel is a waveguide with rough walls, moving machinery and no line of sight to anywhere useful, and a collapsed building is worse. These are the cases where standard models fail and where getting communication to work matters most. The 2009 survey of underground mine communications came out of this, and it remains the most cited thing I have written.
Cognitive radio and spectrum awareness
If a radio can observe what the spectrum is doing, it can predict and adapt rather than simply transmit and hope. That covers spectrum occupancy prediction from past observation, and using what a device knows about its own location and surroundings to make better decisions — the subject of two granted or filed US patents.
Interference management and cross-layer adaptation
Scheduling, power control and adaptation decisions that cut across the layer boundaries they are usually designed within. Inter-cell interference in OFDMA systems is the concrete case.
Selected work
- Underground mine communications: a survey. , 2009
- Binary time series approach to spectrum prediction for cognitive radio. , 2007
- Statistical modeling of propagation channels for terahertz band. , 2017
- System and method of exploiting location awareness to improve wireless cognitive radio. , 2012
- Method for scheduling to reduce inter-cell interference for voice communication in OFDMA. , 2010
The full list is on Google Scholar.