Abstract:Motivated by the application of point-to-point communication networks and biological storage, we investigate new channel coding bounds for noisy permutation channels with strictly positive and full-rank square matrices. Our new achievability bounds use $\epsilon$-packing with Kullback-Leibler divergence as a metric to bound the distance between distributions and are tighter than existing bounds. Additionally, Gaussian approximations of achievability bounds are derived, and the numerical evaluation shows the precision of the approximation.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to establish new lower bounds for channel coding in noisy permutation channels. Specifically, the authors focus on noisy permutation channels with strictly positive - definite and full - rank square matrices, and propose new achievability bounds. These bounds use the KL - divergence as a metric to limit the distance between distributions. In addition, the achievability bounds in the form of Gaussian approximation are also derived, and the accuracy of this approximation is demonstrated through numerical evaluation.
### Research Background and Motivation
1. **Out - of - order Arrival Problem in Communication Networks**: In point - to - point communication networks, data packets may arrive at the receiving end with different delays, resulting in out - of - order arrival. This phenomenon is very common in many practical applications, such as Internet communication and biological storage systems.
2. **Biological Storage Systems**: DNA storage systems encode data into unordered codewords of nucleotides, and randomly arranged DNA fragments can be described by noisy shuffling channels.
3. **Deficiencies in Existing Research**: Traditional asymptotic analysis assumptions are not applicable to the case of short code lengths, and the code lengths in practical applications are usually between several hundred and several thousand. Therefore, a more refined analysis for finite code lengths is required.
### Main Contributions
1. **New Achievability Bounds**: The authors propose an ε - packing method based on KL - divergence to construct message sets, and use this method to obtain new achievability bounds that are tighter than the existing bounds.
2. **Gaussian Approximation**: For a given code length and average error probability, the authors give a Gaussian approximation expression for the maximum achievable code length, which has the form:
\[
\log M\approx\ell\log\left(\sqrt{n}-\ell\Phi^{- 1}\left(\frac{\epsilon}{|R|}\right)\right)+\log\lambda,
\]
where $\xi = 2\binom{|Y|}{2}$, $\ell=\text{rank}(W)-1$, $\lambda$ is the volume ratio, and $\theta$ is a constant.
3. **Numerical Verification**: Numerical evaluation shows that the proposed bounds also have high accuracy in the case of short code lengths.
### Key Technologies and Methods
- **KL - Divergence and ε - packing**: Used to measure the distance between distributions and construct message sets.
- **Gaussian Approximation**: Used to analyze the performance in the case of large samples and provides a simple approximation method.
- **ML Decoder**: The maximum - likelihood decoder is used to reduce the number of error events.
### Application Prospects
This research not only provides more accurate theoretical analysis tools for noisy permutation channels, but also provides guidance for the design of actual communication systems, especially in the case of short code lengths, which helps to improve the reliability and efficiency of the system.