Research Summary

Currently, I am a Lecturer in Information Technology at the University of New England and a Research Affiliate at the University of Sydney. From November 2018 to January 2024, I was working as a Postdoctoral Research Associate in Telecommunications at the University of Sydney. I received B.Sc., M.Sc. and PhD degrees from the Peter the Great St. Petersburg Polytechnic University (SPbPU) in 2010, 2012 and 2015, respectively. PhD thesis: Methods of design and decoding of polar codes.

I have substantial experience in Coding Theory and Artificial Intelligence applied to Wireless Communications, which is confirmed by the list of my publications. My research focuses on the design and analysis of error-correcting codes, which are also known as channel codes. The proposed solutions rely on linear algebra, probability theory, graphs, optimization theory, information theory and artificial intelligence. My publications are devoted primarily to polar codes, but I was also working on Reed-Solomon, BCH and generalized concatenated codes. In addition, I have experience with multiple-input multiple-output (MIMO) detection. During the last few years, the effort has been focused on the design of polar codes with high correction capability and acceptable decoding complexity. My current research project develops machine learning-assisted polar code design to ensure reliable data transmission in the next generation of mobile communications (6G).


The proposed algorithms were analyzed analytically, as well as by simulations. For that they were implemented in C/C++ programming language. AWGN channel with BPSK/QAM modulation was considered. The proposed polar subcodes were shown to outperform turbo-codes from LTE and LDPC from WiMAX in terms of decoding error probability and decoding complexity (in the case of code length up to 4096). The proposed SCL-based sequential decoding algorithm for polar codes has very low time complexity. The proposed shortened/punctured polar code design provides codes of arbitrary length with a minimized estimate of the SC decoding error probability. The proposed precoded polar codes of lengths 32 and 64 have the same error-correction capability as the corresponding extended BCH codes but lower decoding complexity under SCL-based decoding. The proposed recursive precoded polar codes of lengths 128 and 256 outperform 5G New Radio polar codes with CRC-11 on average by 0.2 dB. The proposed method for computing the partial weight distribution of punctured, shortened, precoded polar codes is able to explicitly enumerate all codewords up to a certain weight. 


ANZSRC 2020 Field of Research (FoR) Codes: 


My CV

My video presentation "Precoded Polar Codes With High Error-Correction Capability" 

Links: