Channel Characterisation and Modelling for Large MIMO Systems

The employment of multiple antennas at both the transmitter and receiver enables the so-called Multiple-Input Multiple-Output (MIMO) technologies to greatly improve the link reliability and increase the overall system capacity. MIMO has been used in various wireless communication standards, such as the third generation (3G) and fourth generation (4G) wireless systems. While most current MIMO systems utilise 2-8 antennas, the recently proposed large MIMO (or massive MIMO) systems aim to exploit the potentially large capacity gains that would arise in larger arrays of antennas (with tens or hundreds of antennas at both the transmitter and receiver). Large MIMO systems provide a plethora of advantages over conventional MIMO systems, such as higher data rates, better link reliability, and better tradeoff between spectral efficiency and energy efficiency. In large MIMO systems, energy transmission can be optimised by exploiting the many degrees of freedom offered by the many antenna elements, while random impairments, such as thermal noise and co-channel interference, can be averaged out.

For the practical deployment of large MIMO systems, many research challenges need to be addressed. For example, how can we squeeze a large number of antennas into a limited area/volume while still maintaining the low correlations? Another technical challenge in developing large MIMO systems is signal processing complexity. Since by definition a large spatial array is being exploited, transmit and receive signals are quite lengthy thus the search algorithms must be performed over many possible permutations of symbols. The existing search algorithms, such as linear ascent search (LAS) and random-restart reactive Tabu search (RTS) algorithms, assume that the channel has been perfectly estimated, which, given the size of the channel matrix and thus amount of channels to be tracked, seems like an unreasonable assumption.

For the design, performance evaluation, and optimisation of large MIMO systems, proper characterisation and modelling of the underlying channels are indispensable. Since the quantity of channels involved is extremely large, many channels must be estimated accurately and efficiently in a reasonable amount of time. In the current literature, uncorrelated Rayleigh fading channel models have been mostly used while Kronecker based stochastic models were used only in a few cases. These channel models used for the performance evaluation of large MIMO systems are not realistic, implying most probably too optimistic performance evaluation results of the presented algorithms. This project aims to investigate the statistical channel properties, in particular spatial-temporal-frequency correlation properties, and develop realistic channel models for large MIMO systems considering the tradeoff between the model accuracy and complexity.

The requirements for the applicants are as follows:
1. Excellent first degree (first class or good upper second class), or preferably master degree, in electrical engineering, communications engineering, mathematics, physics, or any other relevant scientific/engineering discipline.
2. The applicant should have excellent background knowledge in Wireless Communications, in particular channel modelling. Preferably, the applicant is expected to possess good knowledge of mathematics, statistics, signals and systems, communication theory, and signal processing. Having expertise in Matlab is also advantageous.

For further information, contact Prof C X Wang ([Email Address Removed])