[Paper Review] Enabling Quality-Driven Scalable Video Transmission over Multi-User NOMA System

Xiaoda Jiang, Hancheng Lu+1 more|arXiv (Cornell University)|Jan 16, 2018

Advanced Wireless Communication Technologies15 references3 citations

TL;DR

This paper proposes a quality-driven cross-layer framework for scalable video transmission over multi-user NOMA systems, integrating video quality models with physical-layer power allocation to maximize perceived video quality (PSNR). It introduces a user grouping strategy to reduce complexity and develops both a global optimal algorithm using hidden monotonicity and a polynomial-time suboptimal greedy algorithm, achieving significant PSNR gains over OMA and throughput-optimized NOMA schemes.

ABSTRACT

Recently, non-orthogonal multiple access (NOMA) has been proposed to achieve higher spectral efficiency over conventional orthogonal multiple access. Although it has the potential to meet increasing demands of video services, it is still challenging to provide high performance video streaming. In this research, we investigate, for the first time, a multi-user NOMA system design for video transmission. Various NOMA systems have been proposed for data transmission in terms of throughput or reliability. However, the perceived quality, or the quality-of-experience of users, is more critical for video transmission. Based on this observation, we design a quality-driven scalable video transmission framework with cross-layer support for multi-user NOMA. To enable low complexity multi-user NOMA operations, a novel user grouping strategy is proposed. The key features in the proposed framework include the integration of the quality model for encoded video with the physical layer model for NOMA transmission, and the formulation of multi-user NOMA-based video transmission as a quality-driven power allocation problem. As the problem is non-concave, a global optimal algorithm based on the hidden monotonic property and a suboptimal algorithm with polynomial time complexity are developed. Simulation results show that the proposed multi-user NOMA system outperforms existing schemes in various video delivery scenarios.

Motivation & Objective

• Address the lack of quality-aware design in NOMA systems for real video applications, where perceived quality (QoE) is more critical than raw throughput or reliability.
• Overcome the limitation of existing NOMA systems that optimize only for physical layer metrics (e.g., spectral efficiency) without considering application-layer video content characteristics.
• Formulate a cross-layer optimization problem that jointly considers scalable video encoding (e.g., layered structure) and NOMA transmission to maximize end-to-end video quality.
• Design a low-complexity user grouping strategy based on user location, requested video content complexity, and quality requirements to reduce multi-user interference and SIC latency.
• Develop efficient power allocation algorithms that optimize for average PSNR under diverse user quality constraints, even though the problem is non-concave.

Proposed method

• Proposes a novel user grouping strategy that pairs weaker users requesting low-complexity video sequences with stronger users requesting high-complexity sequences to improve spectral efficiency and reduce SIC delay.
• Integrates a semi-analytical PSNR model for scalable video streams with the NOMA physical layer model, enabling cross-layer optimization based on video content importance and channel conditions.
• Reformulates the multi-user NOMA video transmission problem as a non-concave average PSNR maximization problem under quality constraints, capturing the trade-off between user fairness and video quality.
• Leverages the hidden monotonic property of the non-concave problem to design a global optimal algorithm using polyblock outer approximation, ensuring convergence to the global optimum.
• Develops a suboptimal greedy algorithm inspired by the successive interference cancellation (SIC) process, achieving near-optimal performance with polynomial time complexity.
• Employs superposition coding (SC) at the transmitter and successive interference cancellation (SIC) at receivers to enable simultaneous transmission of multiple users over the same resource block.

Experimental results

Research questions

• RQ1 How can NOMA be effectively adapted for scalable video transmission to improve perceived video quality rather than just spectral efficiency?
• RQ2 What user grouping strategy minimizes implementation complexity and maximizes video quality in multi-user NOMA systems with heterogeneous user channel conditions and video content demands?
• RQ3 How can the cross-layer integration of video application-layer characteristics (e.g., layered encoding) and physical-layer transmission (e.g., power domain multiplexing) be modeled to optimize end-to-end quality?
• RQ4 Can a non-concave quality-driven power allocation problem in NOMA be solved efficiently, and what algorithms achieve global or near-global optimality?
• RQ5 What are the performance gains of the proposed framework over conventional OMA-based video delivery and existing NOMA schemes optimized for throughput or fairness?

Key findings

• The proposed scheme achieves significantly higher average PSNR than OMA-TQFE (e.g., 34.40 dB vs. 33.42 dB at 15 dB SNR) and NOMA-MT (e.g., 34.40 dB vs. 32.92 dB at 15 dB SNR), demonstrating superior video quality across all video sequences.
• For weaker users requesting low-complexity videos (e.g., 'Mobile', 'Foreman'), the proposed scheme outperforms NOMA-MT by up to 4.8 dB in PSNR (30.54 dB vs. 29.66 dB), indicating fairness and robustness.
• The WLBH (weaker users request low-complexity content, better users request high-complexity content) grouping strategy achieves the best PSNR performance, outperforming WHBL and WRBR by up to 1.5 dB, validating the importance of content-aware grouping.
• Visual results show that the proposed scheme reconstructs frames with better clarity—especially for low-complexity videos like 'Foreman'—while OMA-TQFE and NOMA-MT exhibit blurring or distortion in background regions.
• The suboptimal greedy algorithm achieves PSNR within 2% of the global optimal solution while running in polynomial time, making it practical for real-time deployment.
• At high SNR, OMA-TQFE outperforms NOMA-MT due to its QoE-centric design, but the proposed scheme maintains superiority across all SNR levels, especially for weaker users.