Two-way wireless video communication using Randomized cooperation, Network Coding and packet level FEC

Xiaozhong Xu, Özgü Alay, Elza Erkip, Yao Wang, Shivendra S. Panwar

TL;DR

This paper introduces a novel two-way wireless video communication system that combines Network Coding (NC), Randomized Distributed Space-time Coding (R-DSTC), and packet-level Forward Error Correction (FEC). The proposed system significantly outperforms direct transmission and existing cooperative methods, especially for users located far apart, by efficiently managing bandwidth, delay, and error resilience.

What Problem Does This Paper Address?

Real-time two-way video communication, like video conferencing, faces significant challenges in wireless networks: it demands high bandwidth, low latency, and resilience to errors. Traditional approaches struggle to meet these demands simultaneously. Existing cooperative communication methods, while promising, often require tight synchronization and feedback, which are difficult to achieve in dynamic wireless environments. This paper aims to bridge this gap by proposing a robust, low-delay, and efficient system that integrates advanced techniques to overcome these limitations.

Methods

The authors propose a system that integrates three key technologies:

  • Network Coding (NC): Specifically, they use XOR-based network coding, where relays combine (XOR) packets received from two users. This allows for more efficient data forwarding compared to simply retransmitting packets.
  • Randomized Distributed Space-time Coding (R-DSTC): This technique allows multiple relays to transmit packets simultaneously without requiring strict synchronization. Unlike traditional DSTC, R-DSTC is decentralized, meaning relays don't need to know about each other, making it more adaptable to changing network conditions.
  • Packet-Level Forward Error Correction (FEC): To combat packet losses without relying on retransmissions (which add delay), the system employs FEC. This adds redundant parity packets, allowing the receiver to reconstruct lost packets.

The system is designed with a one-way delay constraint, and the authors optimize transmission parameters (rates, time allocation) to maximize the average useful video rate for both users. They compare their proposed scheme, "R-DSTC+NC," against two baselines: "Direct Transmission" and "R-DSTC." They also introduce a variation called "R-DSTC-HR" which reorders transmission hops for better power concentration at longer distances.

Datasets

Not discussed in the paper. The authors mention using H.264/AVC reference software JM for PSNR evaluation, but this is a simulation tool, not a dataset in the traditional sense.

Key Insights and Ideas

  • Synergy of NC and R-DSTC: The combination of Network Coding and Randomized Distributed Space-time Coding proves to be highly effective. NC efficiently combines data, while R-DSTC provides a flexible and robust way to transmit this combined data from multiple relays.
  • Decentralization is Key: The use of R-DSTC, which avoids tight synchronization and coordination among relays, is crucial for practical implementation in dynamic wireless environments.
  • FEC for Delay-Sensitive Applications: Packet-level FEC is a viable alternative to retransmissions for real-time applications where delay is critical.
  • Relaying Benefits Increase with Distance and Density: Cooperative relaying, especially with the proposed R-DSTC+NC scheme, offers greater advantages when users are farther apart and the network is more densely populated with nodes.
  • R-DSTC-HR for Long Distances: Reordering transmission hops in R-DSTC can provide significant gains for users separated by large distances by focusing relay power.

Key Results

  • The proposed "R-DSTC+NC" scheme significantly outperforms both "Direct Transmission" and "R-DSTC" in terms of achievable average video rates.
  • At a user distance of 100 meters, R-DSTC+NC shows gains close to 4/3 over R-DSTC-HR, and these gains are more pronounced at larger distances.
  • In terms of video quality (measured by PSNR), R-DSTC-HR provides considerable improvement over direct transmission. R-DSTC+NC further enhances this by an additional 1-2 dB, leading to significantly better visual quality, especially at longer distances where direct transmission results in poor perceptual quality.
  • The system is most effective when the distance between users is large.

Limitations and Future Work

  • Limitations:
    • The paper assumes a fixed packet size and a perfect code for FEC, which might not hold in all practical scenarios.
  • The analysis relies on simulations and numerical searches for parameter optimization.
  • The impact of interference from other networks or users is not explicitly detailed.
  • Future Work:
    • Extend the work to next-generation cellular networks like 3GPP LTE and IEEE 802.16.
  • Utilize tractable stochastic geometry models to better capture the effects of base station locations and user distributions in future network designs.
  • Investigate the performance in more complex network topologies and with varying interference levels.

Summary updated

Abstract

Two-way real-time video communication in wireless networks requires high bandwidth, low delay and error resiliency. This paper addresses these demands by proposing a system with the integration of Network Coding (NC), user cooperation using Randomized Distributed Space-time Coding (R-DSTC) and packet level Forward Error Correction (FEC) under a one-way delay constraint. Simulation results show that the proposed scheme significantly outperforms both conventional direct transmission as well as R-DSTC based two-way cooperative transmission, and is most effective when the distance between the users is large.
BibTeX
@misc{xu2012two,
  author = {Xiaozhong Xu and Özgü Alay and Elza Erkip and Yao Wang and Shivendra S. Panwar},
  title = {{Two-way wireless video communication using Randomized cooperation, Network Coding and packet level FEC}},
  year = {2012},
  eprint = {1205.3252},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1205.3252},
}

Loading related content...

Figure