Research Projects (Scholarships and funding for trips are readily available. Please email me if you are interested)

Precoding and detection algorithms for multiuser massive MIMO systems

This project will investigate innovative precoding and detection techniques for multiuser massive MIMO systems that will play a big role in future wireless communication networks including satellite systems, wireless local area and cellular networks. The use of very large antenna arrays poses a major challenge to system designers and it is of fundamental importance to investigate ways of designing precoding and detection techniques for this purpose. We will also consider configurations such as cell-free and networked MIMO as well as reconfigurable intelligent surfaces in our studies. In particular, we will focus on the development of scalable algorithms and the analysis of the diversity order and sum rate of the studied techniques. The research activities will be based on the development of a system model, the derivation of algorithms, and the building of simulations and analytical tools.

Signal processing algorithms for data science applications

This project will investigate signal processing algorithms for data science applications such as feature extraction and anomaly detection in sensor systems and communication networks. In particular, we will examine adaptive algorithms that can learn data features, exploit low-rank and sparse properties of the data as well as centralized and distributed processing strategies. Real-word data that includes data from large-scale sensor systems and communication networks will be considered for testing the developed algorithms. The research activities will be based on the development of system models with linear algebra, simulations tools and analytical development.

Super-resolution techniques for multiuser MIMO systems

This project will investigate super-resolution techniques for multiuser MIMO systems. The use of very large antenna arrays leads to high costs of implementation and energy consumption. Therefore, it is of fundamental importance to investigate ways of designing MIMO systems that can operate with modest number of antenna elements and still obtain significant gains in terms of achievable sum rates and diversity order. In particular, we will focus on the development of techniques that can exploit the virtualization of sparse arrays and the use of electromagnetic theory to produce super-resolution effects with MIMO antenna arrays. We will then design precoders and detectors with super-resolution and carry out theoretical analysis showing some fundamental limits. The research activities will be based on the development of a system model, the derivation of algorithms, and the building of simulations and analytical tools.

Distributed communications and signal processing algorithms for IoT and wireless sensor networks

This project will investigate novel distributed algorithms for power control, cooperation and interference cancellation using spread spectrum techniques in Internet of Things (IoT) and wireless sensor networks. The goal is to devise low-complexity and effective algorithms for increasing the capacity and the reliability of these networks. The activities will involve the development of system models, simulation tools and analytical approaches.

Cooperative relaying and resource allocation techniques for wireless networks

Recently, cooperative communications were used to increase the capacity and the reliability of wireless networks by exploiting a novel form of diversity via cooperation. This project will examine novel cooperative diversity techniques in conjunction with resource allocation algorithms for wireless networks. In particular, we will consider narrowband and OFDM systems and will investigate novel distributed space-time/frequency coding, cloud-aided and buffer-aided techniques, physical-layer network coding, resource allocation and relay selection algorithms for improving the performance and the capacity of wireless networks. The activities will be based on mathematical formulation, simulation and analytical tools.

Channel coding techniques and applications for 5G and beyond

In this research project, we will investigate novel encoding and iterative decoding techniques for low-density parity-check (LDPC) codes and polar codes. We will examine novel forms of irregular encoding and more efficient iterative decoding algorithms such as improved versions of the belief propagation and list decoding. Applications in wireless networks including multi-antenna systems and multicarrier communications will be considered along with LDPC and polar code design and innovative decoding strategies. The research activities will be based on mathematical modeling, and the building of simulation and analytical tools.

Low-complexity channel estimation and equalization for OFDM systems

This project will investigate advanced adaptive channel estimation techniques and innovative equalization concepts for OFDM systems in time-varying scenarios. We will examine strategies to model time-varying channels with basis expansion models and techniques to mitigate the inter-carrier interference that arises due to channel variations within an OFDM block. The main applications will be 5G and beyond systems, and DVB systems. The research activities will be based on the formulation of system and data models with linear algebra, simulations tools and analytical development and analysis.

Bit-interleaved coded modulation (BICM) and iterative processing techniques for wireless networks

This project will investigate novel concepts of BICM and iterative processing techniques for wireless networks such as 5G and future systems. We will investigate appropriate mappings and interleaving strategies for BICM schemes, use of side information and innovative code designs. The proposed techniques will be considered in scenarios with relaying, block fading channels and MIMO systems. The research activities will consider the development of a system and data model, the building of simulations and analytical tools.

Joint iterative interference cancellation, data detection and decoding techniques with network MIMO and cell-free systems

This project will investigate novel concepts of joint iterative interference cancellation, data estimation and decoding with network MIMO and cell-free concepts in future wireless systems. The main idea is to formulate the problem of interference cancellation, parameter estimation and decoding as a joint optimisation problem. We will devise novel cost-effective algorithms for implementing the proposed approach in the uplink of MIMO networks. One significant challenge is how to estimate the channel of co-channel users and we will examine novel ways of determining these parameters. We will then apply the novel algorithms to MIMO systems with multiple cells and cell-free deployments, and evaluate the performance of the proposed algorithms against the best methods available. The research activities will be based on the development of a system and data model, the building of simulations and analytical tools.

Adaptive learning algorithms exploiting prior knowledge and applications

This project will investigate innovative methods of adaptive learning for modeling both linear and nonlinear problems that exploit prior knowledge and consider their application to problems in communications, sensor and electronic systems. The activities will involve the use of machine learning techniques, low-rank decompositions, optimization tools and matrix computations. The work will involve the development of system models using linear algebra, simulation tools with MATLAB, FPGA and analytical approaches.

Compressive sensing algorithms using subspace methods

There has a growing recent interest in compressive sensing techniques for solving numerous problems in communications, signal processing, radar and sonar systems. In fact, compressive sensing techniques are important mathematical tools that allow the solution of problems with increased accuracy and lower computational complexity. In this project, we will investigate advanced subspace tracking algorithms and iterative thresholding methods with multipass strategies for solving problems that arise in a variety of applications such as system identification, channel estimation in wireless communications, image deblurring and filtering problems. The main goal is to devise low-complexity and effective algorithms with increased accuracy and low reconstruction errors. The activities will involve the development of system models using linear algebra, simulation tools and analytical approaches.

Robust and low-complexity beamforming algorithms

This project will investigate robust adaptive beamforming algorithms and low-complexity strategies for implementing them in applications of sensing and wireless communications. We will consider both centralized and distributed scenarios along with realistic modeling methods for the sensor arrays. The activities will involve the use of constrained adaptive algorithms, low-rank decompositions, optimization tools and matrix computations. The work will involve the development of system models using linear algebra, simulation tools with MATLAB, FPGA and analytical approaches.

High-resolution direction finding algorithms

This project will investigate direction finding algorithms and low-complexity strategies for implementing them in applications of sensing, localisation and wireless communications. We will consider both centralized, sparse and distributed scenarios along with realistic modeling methods for the sensor arrays. The activities will involve the use of subspace tracking algorithms, MUSIC and ESPRIT methods, super-resolution techniques with sparse arrays, optimization tools and matrix computations. The work will involve the development of system models using linear algebra, simulation tools with MATLAB, FPGA and analytical approaches.

Space-time processing algorithms for radar and sonar systems

This project will investigate a novel joint space-time processor for radar and sonar systems. We will investigate novel reduced-rank signal processing algorithms for multidimensional data and the use of prior knowledge for devising high performance target detection algorithms. The research activities will be based on the development of a signal model, simulations tools and analytical development and analysis.

Kernel-based adaptive signal processing algorithms and applications

This project will investigate signal modeling problems that arise in the design power amplifiers and time series with the use of kernel-based adaptive signal processing algorithms. An investigation into variable structures and low-rank techniques using kernels will be carried out. We will examine novel kernel-based adaptive signal processing algorithms with attractive tradeoffs between performance and complexity for modeling and learning. The research activities will be based on the development of system models with linear algebra, simulations tools and analytical development.