Thesis background and motivation
For future communication networks, the demand for higher spectral efficiency while keeping power consumption and hardware/computational complexity at acceptable levels calls for novel architectural and algorithmic solutions. On the network architecture side, alternative to cellular MIMO (CMIMO), distributed MIMO (DMIMO) is seen as a promising candidate where enhanced signal power and/or interference management can be achieved by various modes of base-station cooperation e.g. coordinated beamforming (CBF), non-coherent or coherent joint transmission (JT). On the hardware architecture side, as an alternative to fully-digital beamforming (DBF), hybrid analog/digital beamforming (HBF) is a good candidate which can provide desirable tradeoffs between achievable spectral efficiency, power consumption and complexity by having a small number of RF chains.
The works on these two technologies has so far progressed separate from each other, i.e., most prior DMIMO literature considers only DBF architecture whereas most of the works in HBF literature focuses on CMIMO with each base-station independently designing its own precoders using local CSI. Therefore, there is a great research potential in the combined HBF DMIMO system especially in terms of precoder optimization and analysis of achievable performance and complexity trade-off. In this context, preliminary goals and associated tasks of this project can be listed as:
Goal 1: Performance analysis of HBF CMIMO without cooperation which can serve as a baseline
- Building MATLAB simulator (Existing simulators in Huawei Sweden can be utilized.)
- Empirical analysis of key metrics and KPI’s e.g. sum-rate, cell-edge-rate, SINR..etc.
Goal 2: Precoder design for HBF DMIMO with various modes of base-station cooperation that can achieve good performance and complexity trade-offs
- Literature review on HBF precoder design to determine appropriate optimization algorithms
- Optimization of analog and digital precoders with practical system constraints
- Testing the designed precoders under various scenarios of interest
- Master student in Electrical Engineering, Computer Science or equivalent.
- Experience in modeling, link level simulation and MATLAB.
- A solid theoretical background in communication theory, signal processing and machine learning.
Start: November 2021
Number of positions: 1