PHY2 – Performance evaluation
Tuesday, 4 June 2024, 16:00-17:30, room Pelican
Session Chair: Sofie Pollin (KU Leuven, BE)
On the Optimality of Support Vector Machines for Channel Decoding
Gastón De Boni Rovella (ISAE-SUPAERO, France); Meryem Benammar (ISAE-Supaero, France); Hugo Méric (CNES, France); Tarik Benaddi (Thales Alenia Space, France)
In this work, we investigate the construction of channel decoders based on machine learning solutions, and more specifically, Support Vector Machines (SVM). The channel decoding problem being a high-dimensional multiclass classification problem, previous attempts were made in the literature to construct SVM-based channel decoders. However, existing solutions suffer from a dimensionality curse, both in the number of SVMs involved –which are exponential in the block length– and in the training dataset size. In this work, we revisit SVM-based channel decoders by alleviating these limitations and prove that the suggested SVM construction can achieve optimal Bit Error Probability (BEP) by attaining the performance of the bit-Maximum A Posteriori (MAP) decoder in the Additive White Gaussian Noise (AWGN) channel.
NOMA Performance Subject to Mobility Constraints: A 5G Link-Level Simulator
Md Shantanu Islam, John Hayes, Raouf Abozariba and Adel Aneiba (Birmingham City University, United Kingdom (Great Britain)); Taufiq Asyhari (Monash University, Indonesia)
Analyzing multilayer Non-Orthogonal Multiple Access (NOMA) under mobility is a complex task that is not sufficiently treated in the literature. This paper presents a link-level Simulator compliant with 3GPP Fifth-Generation (5G) physical-layer technologies with NOMA downlink extension, capable of performing detailed mobility analysis. The simulator evaluates rural and urban environments by implementing standardized channel profiles and mobility models such as Random Waypoint (RWP) and Manhattan. We use the dynamic capabilities of the simulator to assess the link sustainability of heterogeneous mobile NOMA users under multiple antenna systems (MIMO) and various beamwidth configurations. The simulator provides a mobility environment for benchmarking and performance evaluation of multilayer NOMA technology on standard 5G New Radio (NR) transceivers.
Feasibility of DFT-s-OFDM in 6G Downlink: Coverage and Energy-Efficiency Assessment
Lauri Anttila and Vili Toivonen (Tampere University, Finland); Oskari Tervo and Esa Tiirola (Nokia Bell Labs, Finland); Kari Hooli (Nokia, Finland); Jukka Talvitie and Mikko Valkama (Tampere University, Finland)
The future 6G mobile communication systems are expected not only to improve the broadband connectivity and other services but also enhance the network coverage, energy-efficiency and sustainability. In this paper, we address the coverage and energy-efficiency challenges, and study the feasibility of discrete Fourier transform spread orthogonal frequency division multiplexing (DFT-s-OFDM) as the potential downlink physical-layer radio access waveform. Building on the prevailing 5G New Radio (NR) downlink transmit signal quality requirements and behavior of contemporary power amplifier (PA) systems, we assess and show the available downlink coverage gain and energy saving potential, compared to ordinary cyclic prefix (CP) OFDM. The obtained results show that DFT-s-OFDM allows improving the downlink coverage by up to 20%, with largest gains available for low modulation orders. The results also show that truly saving energy for a given throughput requirement is more difficult – yet possible, assuming that the PA bias and supply voltages can be properly controlled during the downlink transmission.
Impact of Channel Estimation Feedback on 5G NR Integrated Access and Backhaul
Khaled Tahkoubit (CEA-Leti, France); Nicolas Cassiau (CEA-Leti Minatec Campus, France); Rafik Zayani and Yaya Bello (CEA-LETI, France)
Integrated Access and Backhaul (IAB) systems are gaining more and more interest for future wireless networks. IAB combines the access and the backhaul links to improve network efficiency and reduce deployment costs. This study investigates the impact of the channel estimation feedback on digital precoding in 5G NR Mixed Numerology IAB systems. Combining analog beamforming with Reconfigurable Intelligent Surface (RIS) and digital precoding, we address challenges related to inter-user and inter-numerology interference. We rely on the model of an existing RIS and on realistic channel modeling. We demonstrate the effectiveness of hybrid precoding in interference mitigation, enhancing system performance with limited complexity. Simulation results affirm the potential of our approach for high-capacity and interference-resilient wireless networks, offering valuable insights for future designs.
Secrecy Energy-Efficient Multi-User NOMA: Closed-Form Solution to Bi-Criterion Formulation
Anne Savard, Giulia Cervia, Miled Alam and Arthur Louchart (IMT Nord Europe, France)
This paper investigates the secrecy energy-efficiency (SEE) of a multi-user downlink non-orthogonal multiple access (NOMA) transmission in the presence of a passive eavesdropper. The SEE, capturing the trade-off between the achievable secrecy sum-rate and the power consumption, is here formulated as a bi-criterion optimization problem that is shown to be convex. As such, the characterization of the Pareto-boundary boils down to maximizing the scalarized trade-off between the secrecy sum-rate and the power consumption, which is solved in closed-form. Finally, this obtained closed-form is exploited to maximize the ratio between the secrecy sum-rate and the power consumption. Our numerical results highlight a significant relative SEE gain between NOMA and OMA.
