Placeholder image for staff profiles

Dr Chathura Jayawardena


My publications

Publications

Nikitopoulos Konstantinos, Georgis Georgios, Jayawardena Chathura, Chatzipanagiotis Daniil, Tafazolli Rahim (2018) Massively Parallel Tree Search for High-Dimensional Sphere Decoders,Transactions on Parallel and Distributed Systems IEEE
The recent paradigm shift towards the transmission of large numbers of mutually interfering information streams, as in the
case of aggressive spatial multiplexing, combined with requirements towards very low processing latency despite the frequency
plateauing of traditional processors, initiates a need to revisit the fundamental maximum-likelihood (ML) and, consequently, the
sphere-decoding (SD) detection problem. This work presents the design and VLSI architecture of MultiSphere; the first method to
massively parallelize the tree search of large sphere decoders in a nearly-concurrent manner, without compromising their
maximum-likelihood performance, and by keeping the overall processing complexity comparable to that of highly-optimized sequential
sphere decoders. For a 10 å 10 MIMO spatially multiplexed system with 16-QAM modulation and 32 processing elements, our
MultiSphere architecture can reduce latency by 29å against well-known sequential SDs, approaching the processing latency of linear
detection methods, without compromising ML optimality. In MIMO multicarrier systems targeting exact ML decoding, MultiSphere
achieves processing latency and hardware efficiency that are orders of magnitude improved compared to approaches employing one
SD per subcarrier. In addition, for 16å16 both ?hard?- and ?soft?-output MIMO systems, approximate MultiSphere versions are shown to
achieve similar error rate performance with state-of-the art approximate SDs having akin parallelization properties, by using only one
tenth of the processing elements, and to achieve up to approximately 9å increased energy efficiency.
Jayawardena Chathura, Nikitopoulos Konstantinos Massively Parallel Detection for Non-Orthogonal Signal Transmissions,Proceedings of the IEEE GLOBECOM 2018 Workshops Institute of Electrical and Electronics Engineers (IEEE)
The increasing demand for massive connectivity
with low latency requirements has triggered a paradigm shift
towards Non-Orthogonal transmissions. Still, to translate the
theoretical gains of Non-Orthogonal transmissions into practical,
efficient ?soft? detection schemes are required. The detection
latency and/or complexity of state-of-the-art detection methods
becomes impractical for large Non-Orthogonal systems, both
due to the large number of interfering streams and due to the
rank-deficient or ill-determined nature of the corresponding interference
matrix. Extending the recently proposed MultiSphere
framework, this work introduces NorthCore; a massively parallel
sphere-decoding-based scheme for the detection of large and illdetermined
Non-Orthogonal systems. Similarly to MultiSphere,
NorthCore reduces the corresponding search space by focusing
the available processing power to the most promising vector
solutions that are processed in parallel. As a result, the proposed
detection scheme can attain a detection processing latency similar
to that of highly-suboptimal linear detectors and even outperform
state-of-the-art sophisticated detection approaches with up to
an order of magnitude reduced complexity. To identify the
most promising vector solutions, NorthCore introduces a sortfree
candidate selection technique that reduces the necessary
preprocessing complexity by up to an order of magnitude, making
the proposed approach practical.
Jayawardena Chathura, Nikitopoulos Konstantinos (2019) G-MultiSphere: Generalizing Massively Parallel Detection for Non-Orthogonal Signal Transmissions,IEEE Transactions on Communications pp. 1-12 Institute of Electrical and Electronics Engineers (IEEE)
The increasing demand for connectivity and throughput, despite the spectrum limitations, has triggered a paradigm shift towards non-orthogonal signal transmissions. However, the complexity requirements of near-optimal detection methods for such systems becomes impractical, due to the large number of mutually interfering streams and to the rank-deficient or ill-determined nature of the corresponding interference matrix. This work introduces g-MultiSphere; a generic massively parallel and near-optimal sphere-decoding-based approach that, in contrast to prior work, applies to both well- and ill-determined non-orthogonal systems. We show that g-MultiSphere is the first approach that can support large uplink multi-user MIMO systems with numbers of concurrently transmitting users that exceed the number of receive antennas by a factor of two or more, while attaining throughput gains of up to 60% and with reduced complexity requirements in comparison to known approaches. By eliminating the need for sparse signal transmissions for nonorthogonal multiple access (NOMA) schemes, g-MultiSphere can support more users than existing systems with better detection performance and practical complexity requirements. In comparison to state- of-the-art detectors for NOMA schemes and nonorthogonal signal waveforms (e.g., SEFDM) g-MultiSphere can be up to an order of magnitude less complex, and can provide throughput gains of up to 60%.
The increasing demand for connectivity and throughput, combined with the tight latency requirements of current communication systems, and the existing spectrum limitations, has triggered a paradigm shift towards non-orthogonal signal transmissions where multiple information streams are transmitted using the same time/frequency resources. Despite the promising theoretical gains of such transmissions, the complexity and/or latency requirements of the corresponding receiver processing techniques that are required to translate these gains into throughput make their realization impractical, especially for large numbers of mutually interfering information streams. In addition to the processing complexity/latency increase related to systems? high dimensionality, the interference matrix of some of the recently proposed non-orthogonal transmission schemes can either be ill-determined or even rank-deficient, making their detection even more challenging. These requirements combined with the saturating speed of processors, motivates a timely requirement for massively parallel processing of such non-orthogonal transmissions. In this context, this thesis introduces a generic, massively parallel and near-optimal processing framework that applies to both well- and ill-determined non-orthogonal systems. In contrast to known approaches, this framework enables practical large uplink multi-user MIMO systems with numbers of concurrently transmitting users that exceed the number of receive antennas by a factor of two or more. In contrast to traditional approaches, the proposed framework, does not require sparse signal transmissions for the detection of non-orthogonal multiple access (NOMA) schemes since it is not based on the ?Message Passing? algorithm. Consequently, the proposed framework can enable more efficient NOMA approaches that support more users than existing systems, with better detection performance and practical complexity requirements. In comparison to state-of-the-art detectors for NOMA schemes and non-orthogonal signal waveforms(e.g., Spectrally efficient FDM) the proposed scheme can be up to an order of magnitude less complex and can provide throughput gains of up to 60%.This thesis also introduces a massively parallel soft-input soft-output (SISO) detection design for large MIMO systems capable of bridging the gap between theoretical capacity and achievable throughput, with a processing complexity that can be an order of magnitude lower than that of highly optimized sequential SISO detectors, and a processing latency similar to that of highly sub-optimal, linear, SISO detection approaches. Finally, a massively parallel processing framework is presented that enables extreme grant-free non-orthogonal multiple access. This framework allows reliable and low-overhead user identification and reliable detection/decoding with complexity requirements that can be orders of magnitude lower than existing schemes.