[Paper Review] Anticipatory Buffer Control and Resource Allocation for Wireless Video Streaming

Sanam Sadr, Stefan Valentin|arXiv (Cornell University)|Apr 10, 2013

Advanced Wireless Network Optimization6 references21 citations

TL;DR

This paper proposes an anticipatory buffer control and resource allocation scheme for wireless video streaming that predicts future channel conditions to pre-load buffers during good channel states, minimizing bandwidth usage while ensuring smooth playback. By formulating the problem as a linear program, the scheme optimally trades off buffer size and spectrum allocation, enabling support for more users at high video quality even under poor future channel conditions.

ABSTRACT

This paper describes a new approach for allocating resources to video streaming traffic. Assuming that the future channel state can be predicted for a certain time, we minimize the fraction of the bandwidth consumed for smooth streaming by jointly allocating wireless channel resources and play-out buffer size. To formalize this idea, we introduce a new model to capture the dynamic of a video streaming buffer and the allocated spectrum in an optimization problem. The result is a Linear Program that allows to trade off buffer size and allocated bandwidth. Based on this tractable model, our simulation results show that anticipating poor channel states and pre-loading the buffer accordingly allows to serve more users at perfect video quality.

Motivation & Objective

• To address the challenge of maintaining smooth video streaming in cellular networks with limited radio resources, especially at cell edges or during handover.
• To minimize the bandwidth consumption for smooth video streaming by jointly optimizing buffer size and data rate allocation.
• To leverage predicted future channel states to proactively allocate resources before poor channel conditions occur.
• To enable a trade-off between buffer size and spectrum usage while ensuring uninterrupted playback.
• To design a low-complexity, scalable solution compatible with OFDMA-based systems like LTE.

Proposed method

• The authors model the video buffer dynamics using a linear programming framework that tracks allocated, played, and remaining bits over time.
• A look-ahead window of T time slots is used to predict average channel gain, enabling proactive resource allocation.
• The optimization problem balances buffer fullness and spectrum efficiency, ensuring the minimum required data rate is met at all times.
• The scheme pre-loads the buffer during good channel conditions when poor future conditions are predicted, reducing reliance on real-time channel quality.
• Resource allocation is adjusted based on predicted channel gains, with reduced data rates at cell edges and increased pre-loading during favorable conditions.
• The model is independent of video codec and subjective quality metrics, focusing on objective bit rate constraints.

Experimental results

Research questions

• RQ1 How can future channel state predictions be used to improve resource allocation for video streaming?
• RQ2 What is the optimal trade-off between buffer size and allocated bandwidth to ensure smooth playback with minimal spectrum usage?
• RQ3 How does anticipatory pre-loading reduce the risk of playback outages during poor channel conditions?
• RQ4 What is the impact of buffer pre-loading on system capacity and the number of supported video users?
• RQ5 How does the proposed scheme compare to non-anticipatory approaches in terms of bandwidth efficiency and user throughput?

Key findings

• The proposed scheme supports a higher number of video users compared to non-anticipatory systems, especially under high load or poor channel conditions.
• By pre-loading the buffer during good channel states, the system reduces the need for high data rates during poor conditions, saving spectrum.
• The linear programming formulation enables fast and scalable implementation with low computational complexity.
• With a maximum buffer size of Z = 5V, the scheme maintains smooth playback even when channel quality degrades at the cell edge.
• The system achieves 100% service rate (all users receive required bit rate) under optimal conditions, demonstrating high efficiency.
• The scheme effectively prevents playback outages by proactively managing buffer levels based on predicted channel states.