Joint Source, Channel and Space-time Coding of Progressive Sources in MIMO Systems

Sang-Hyo KimarXiv 2016 · arXiv

TL;DR

This paper tackles the complex problem of jointly optimizing source, channel, and space-time coding for progressive multimedia sources in MIMO systems. By introducing a parametric distortion-rate model, the authors develop an efficient packet-by-packet optimization method that significantly reduces computational complexity while improving transmission performance.

What Problem Does This Paper Address?

Transmitting progressive multimedia content (like scalable video or embedded images) over wireless channels is challenging due to the inherent unreliability of these channels. Progressive sources allow for graceful degradation of quality as packet loss increases, but their bitstreams are sensitive to errors. In modern wireless systems employing Multiple-Input Multiple-Output (MIMO) technology, the complexity escalates dramatically. This is because not only do we need to optimize how source data is encoded (source coding) and protected against errors (channel coding), but we also need to decide how to utilize the multiple antennas for diversity and multiplexing gains (space-time coding). The sheer number of combinations for assigning channel codes, modulations, and space-time codes to each packet of a progressive stream becomes computationally prohibitive for exhaustive search. Existing methods for simpler SISO systems don't directly extend to the MIMO scenario with space-time coding.

Methods

The core of this paper's contribution lies in its parametric approach to tackle the joint optimization problem. Instead of using the actual, complex distortion-rate function of a source coder, they introduce a parametric distortion-rate function ($f^P(x) = \sigma^2 2^{-\alpha x}$). This function, parameterized by $\alpha$, allows for a more tractable mathematical analysis.

The authors then derive theoretical results based on this parametric model, which enable them to propose an efficient optimization method. This method optimizes the assignment of spectral efficiencies (related to channel coding and modulation) and spatial multiplexing rates (related to space-time codes) to progressive packets in a packet-by-packet manner. This sequential optimization significantly reduces the computational complexity compared to a brute-force, exhaustive search over all possible assignments. They also derive constraints on the search space of these parameters, further enhancing efficiency.

Datasets

Not discussed in the paper. The paper focuses on theoretical development and numerical evaluation using simulated data.

Key Insights and Ideas

  • The Power of Parametric Modeling: By abstracting the source's distortion-rate characteristic into a parametric form, the authors unlock a path to efficient optimization that was previously intractable.
  • Packet-by-Packet Optimization: The key to managing complexity in MIMO systems with progressive sources is to break down the problem into smaller, sequential decisions for each packet. This "greedy" approach, guided by theoretical results, proves highly effective.
  • MIMO Complexity vs. SISO Simplicity: The paper highlights how the introduction of space-time coding in MIMO systems fundamentally changes the optimization landscape, making simple SISO strategies insufficient. The interaction between spatial multiplexing and error probability is more complex than in SISO systems.
  • Trade-off Between Distortion and Error Probability: The paper emphasizes that increasing spectral efficiency or spatial multiplexing rate can reduce source distortion but increases packet error probability. The optimization aims to find the best balance.

Key Results

  • The proposed packet-by-packet optimization method significantly reduces computational complexity compared to exhaustive search, from $(N_{se}N_{smr})^L$ to $QN_{se}N_{smr}L$ (where $Q$ is the number of quantization levels for the parameter $\alpha$).
  • Numerical results demonstrate that the proposed method significantly improves the Peak-Signal-to-Noise Ratio (PSNR) performance of a rate-based optimal solution in MIMO systems.
  • In SISO systems (where space-time coding is not used), the proposed method's performance is close to that of state-of-the-art local search algorithms.
  • The performance gains are observed across different MIMO channel conditions (i.i.d. Rayleigh, spatially correlated Rayleigh, and Rician fading).
  • The paper shows that excluding certain space-time codes can lead to a significant PSNR degradation, highlighting the importance of jointly considering them.

Limitations and Future Work

  • Suboptimal Performance: The paper acknowledges that the parametric distortion-based solution is suboptimal compared to an exhaustive search, which would yield the true optimal performance. However, the computational cost of exhaustive search makes it impractical.
  • Parameter $\alpha$ Selection: While the parametric approach simplifies optimization, selecting the "best" parameter $\alpha$ still requires an exhaustive search over a quantized range of $\alpha$ values.
  • Future Work: The authors suggest extending their joint source, channel, and space-time coding framework to progressive transmission in multiview settings, where multiple cameras' streams are coded and sent to clients. This would involve handling even more complex dependencies and optimization challenges.

Summary updated

Abstract

The optimization of joint source and channel coding for a sequence of numerous progressive packets is a challenging problem. Further, the problem becomes more complicated if the space-time coding is also involved with the optimization in a multiple-input multiple-output (MIMO) system. This is because the number of ways of jointly assigning channels codes and space-time codes to progressive packets is much larger than that of solely assigning channel codes to the packets. We are unaware of any feasible and complete solution for such optimization of joint source, channel, and space-time coding of progressive packets. This paper applies a parametric approach to address that complex joint optimization problem in a MIMO system. We use the parametric methodology to derive some useful theoretical results, and then exploit those results to propose an optimization method where the joint assignment of channel codes and space-time codes to the packets can be optimized in a packet-by-packet manner. As a result, the computational complexity of the optimization is exponentially reduced, compared to the conventional exhaustive search. The numerical results show that the proposed method significantly improves the peak-signal-to-noise ratio performance of the rate-based optimal solution in a MIMO system.
BibTeX
@misc{kim2016joint,
  author = {Sang-Hyo Kim},
  title = {{Joint Source, Channel and Space-time Coding of Progressive Sources in MIMO Systems}},
  year = {2016},
  eprint = {1610.04378},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1610.04378},
}

Loading related content...

Figure