QoE-Aware Resource Allocation for Video Transmissions over MIMO Systems

Qian Wang, Yilong Yin, Xiushan Nie2017 · DOI

Abstract

With the rapid increasing demands of wireless media applications, it becomes more and more important on video transmission technologies. With the rapid growth of media application, QoE (quality of experience), that is the standard of quality of service, many researchers focus on the topic of the application layer and physical layer joint optimization.

Introduction

As wireless video consumption continues to dominate network traffic, traditional Quality of Service (QoS) metrics—such as raw bit rate or latency—are increasingly insufficient for measuring actual user satisfaction. A high-throughput connection can still result in a poor viewing experience if the video stutters due to buffer underflows or exhibits high distortion from aggressive compression. To address this, researchers are shifting focus toward Quality of Experience (QoE), which more accurately reflects human perception by accounting for the non-linear relationship between data rates and perceived quality.

Wireless channels, particularly in Multiple-Input Multiple-Output (MIMO) systems, are inherently volatile and prone to interference. Concurrently, video data possesses specific Rate-Distortion (RD) characteristics. This paper proposes a cross-layer strategy that bridges the gap between the application layer (video encoding) and the physical layer (wireless transmission), allowing for a more holistic optimization of the end-to-end video delivery chain.

Methodology

The proposed framework models a MIMO-OFDM (Orthogonal Frequency Division Multiplexing) system where resource allocation is driven by QoE requirements. The methodology integrates three critical components:

  1. QoE Modeling: A logarithmic utility function is employed to represent QoE, capturing the "law of diminishing returns" where incremental bit rate increases yield decreasing marginal improvements in perceived quality.
  2. Video Rate-Distortion (RD) Function: The system incorporates a mathematical model relating video distortion to quantization parameters and transmission rates.
  3. Cross-Layer Optimization: The authors formulate an optimization problem to maximize aggregate QoE across all users, subject to total power constraints and physical subcarrier capacities. Using Karush-Kuhn-Tucker (KKT) conditions, they derive a closed-form solution for optimal quantization parameters ($q_i^*$), enabling real-time adaptation to fluctuating channel conditions.

The system was evaluated using numerical simulations of a MIMO-OFDM environment with 32 subcarriers and a varying number of users (4 to 12). Physical parameters included a path-loss exponent of 2.5 and a noise power density of -150 dB/Hz.

Performance Evaluation

The cross-layer approach demonstrates significant advantages over traditional decoupled methods like Iterative Water-Filling (IWF).

Queue Stability

One of the primary benefits of the proposed scheme is the maintenance of stable aggregate queue backlogs. As shown in Figure 2, the cross-layer allocation prevents the rapid growth of backlogs seen in the IWF baseline, which is prone to bottlenecks and subsequent video buffering.

Aggregate queue backlog comparison between the proposed cross-layer method and the IWF baseline.

Distortion Reduction and Convergence

The system also achieves lower video distortion, measured by the average Mean Squared Error (MSE). The optimization algorithm converges rapidly, with the most substantial quality improvements occurring within the first few iterations. This efficiency is critical for real-time multimedia applications where decisions must be made in milliseconds.

Average MSE vs. Number of Iterations for different user scenarios, showing rapid convergence and lower distortion.

Conclusion

By prioritizing user perception over raw throughput, the proposed QoE-aware resource allocation framework ensures smoother video delivery and more efficient network utilization. The derivation of a closed-form solution for optimal encoding parameters makes this approach computationally feasible for practical deployment in modern base stations. Future work will explore more complex utility functions that account for content-specific characteristics and high-mobility environments.

Summary updated

Abstract

With the rapid increasing demands of wireless media applications, it becomes more and more important on video transmission technologies. With the rapid growth of media application, QoE (quality of experience), that is the standard of quality of service, many researchers focus on the topic of the app
BibTeX
@misc{wang2017qoe,
  author = {Qian Wang and Yilong Yin and Xiushan Nie},
  title = {{QoE-Aware Resource Allocation for Video Transmissions over MIMO Systems}},
  year = {2017},
  doi = {10.4108/eai.13-7-2017.2270678},
}
Figure