PhD - Leuven | More than two weeks ago
You will design the fastest and most energy-efficient wireless communication systems of the future
Over the last decades, wireless communications have undergone an exponential growth, providing an ever-increasing throughput thanks to higher bandwidth, spatial multiplexing, and improved implementation of high-speed communication schemes. Going beyond throughputs offered by 5G networks and the latest WLAN standards such as 802.11ax/ay, future systems will aim at speeds of 100 Gbps or more. This requires disruptive solutions, and one key ingredient is to consider carrier frequencies above 100 GHz where a broad spectrum is available and large multiple-antenna systems can be integrated within small dimensions.
With a bandwidth of 1 to 10 GHz around those frequencies, the complexity of the baseband DSP explodes and requires innovative solutions co-optimizing algorithms and architectures for wideband MIMO and beamforming, equalization, or filtering. This co-design is essential, considering both system performance and implementation complexity in order to identify the best trade-offs, propose reduced-complexity solutions and optimize quantization accuracy.
Beyond high-level algorithmic refinement, designing digital architectures able to cope with bandwidth significantly larger than achievable clock frequencies will require dedicated parallelization and pipelining solutions, to be investigated in this PhD. The interoperation with the analog front-end will also be essential, making sure high-throughput interfacing problems are solved and wideband analog non-idealities can be compensated digitally.
This PhD will build on imec's experience of high-throughput mm-wave communication systems and identify the key components to optimize based on state-of-the-art literature and assessment of complexity and performance of candidate DSP solutions. The work will combine Matlab-based performance simulation with a digital design flow to validate speed and complexity. This research may be combined with experiments and measurement on a communication testbed.
Required background: Electrical Engineering, Digital Signal Processing for Communications, Digital Circuit Design (VHDL or Verilog), Matlab, C/C++
Type of work: 10% literature/theory, 60% modeling/simulation, 30% design/experimental
Supervisor: Sofie Pollin
Daily advisor: Claude Desset
The reference code for this position is 2021-133. Mention this reference code on your application form.