Rate-Adaptive Protograph-Based Raptor-Like LDPC Code for

Continuous-Variable Quantum Key Distribution

Erdem Eray Cil1 and Laurent Schmalen1
1
Communications Engineering Lab, Karlsruhe Institute of Technology
Posted on 6 Aug 2024 — CC-BY-NC-SA 4 — https://doi.org/10.36227/techrxiv.172297422.27913742/v1 — e-Prints posted on TechRxiv are preliminary reports that are not peer reviewed. They should not b...

August 06, 2024

Abstract
We propose a new type-based protograph raptor-like LDPC code for rate-adaptive information reconciliation in CV-QKD
systems. It offers robust error-correction performance across a wide range of rates while simplifying the code design process.

1
 Rate-Adaptive Protograph-Based Raptor-Like LDPC
Code for Continuous-Variable Quantum Key Distribution
Erdem Eray Cil* and Laurent Schmalen
Communications Engineering Lab, Karlsruhe Institute of Technology, Karlsruhe, Germany
* erdem.cil@kit.edu

Abstract: We propose a new type-based protograph raptor-like LDPC code for rate-adaptive
information reconciliation in CV-QKD systems. It offers robust error-correction performance
across a wide range of rates while simplifying the code design process. © 2024 The Author(s)
1. Introduction
Information reconciliation (IR) of continuous-variable quantum key distribution (CV-QKD) systems necessitates
the use of low-rate forward error correction codes that can operate near the capacity limit of the binary-input
AWGN channel. For low to ultra-low rates, multi-edge type (MET) low-density parity check (LDPC) codes are
among the best performing codes [1]. However, optimizing MET LDPC codes is difficult due to the large search
space. To address this, the type-based protograph (TBP) optimization approach is proposed in [2]. This approach
limits the nodes in the protograph to defined types, thereby simplifying the optimization process.
In practical implementations of CV-QKD systems, using a rate-adaptive code is crucial, as real-world applica-
tion parameters are not fixed and fluctuate over time. Hence, to optimize secret key generation, the code rate in
information reconciliation must be adaptable. Rate adaptation can be realized through puncturing or shortening the
code, however, these approaches may degrade error correction performance and require additional signaling [3].
Raptor-like (RL) LDPC codes present an attractive alternative for rate adaptation. Using a selected portion of
the parity check matrix (PCM), rate adaptation can be achieved with simplicity. For example, an RL LDPC code
for CV-QKD systems that operates efficiently within the rate range of [0.016, 0.034] is designed in [4].
Protograph-based raptor-like (PBRL) LDPC codes have been proposed to reduce the design complexity of RL
LDPC codes by imposing constraints on the code graph [5]. These codes provide rate adaptability with exceptional
performance, making them a preferred choice in the current 5G standard [6]. However, for low-rate code designs,
the design complexity of PBRL LDPC codes remains high due to the large size of the protomatrix that needs to
be optimized, in contrast to high-rate code designs.
In this paper, we propose a novel method for designing PBRL LDPC codes [5] using type-based protographs [2].
The proposed type-based protograph raptor-like (TBP-RL) LDPC codes offer robust error correction performance
with low design complexity due to a smaller search space. This simplification enables the design of codes that can
operate across a wide range of rates. To illustrate this, we optimize a code for the wide rate range of [0.01, 0.2]
and evaluate its performance in terms of frame error rate (FER) and asymptotic threshold.
2. Protograph Optimization
The structure of a PBRL LDPC code comprises the highest rate code protograph and an incremental redundancy
code (IRC) protograph [5]. All check nodes within the IRC protograph are connected to a single degree-1 vari-
able node. By incorporating an additional check node into the IRC protograph, the code rate can be reduced.
Consequently, code optimization first involves optimizing the protograph of the highest rate code, followed by
an incremental optimization of the extension checks in the IRC protograph to minimize the asymptotic threshold.
Using type-based optimization [2], the optimization problem is simplified to choosing the type that yields the
minimum asymptotic threshold for each extension check node.
This code optimization process highlights the primary advantage of this code: the simplicity of the rate-
adaptation process. Since the PCM is optimized incrementally for various rates, rate adaptation entails selecting
the appropriate size of PCM. The encoder maps the message to the initial variable nodes based on the desired
code rate, and the decoder determines the code rate from the received codeword length. This eliminates the need
for additional signaling for puncturing/shortening patterns. As a result, the implementation of the IR system is
simplified with RL-LDPC codes, since a single parity-check matrix is sufficient for rate adaptation.
3. Results and Evaluation
To demonstrate the feasibility of the proposed TBP-RL LDPC code design, we first optimize the highest rate code
with a rate of R = 0.2. Subsequently, we incrementally optimize the extension checks to form the IRC protograph,
achieving a minimum rate of R = 0.01. For this, we use the expanded type definitions from [2].
Figure 1 illustrates the asymptotic thresholds of the optimized TBP-RL LDPC code at various rates. The figure
reveals that constraining the search space in TBP-RL LDPC code optimization does not lead to performance losses
when compared to the unrestricted optimization of MET LDPC [1] or TBP LDPC codes [2]. Consequently, the
performance of multiple codes specifically designed for single discrete rates can be achieved using one TBP-RL
 −0.6
Capacity
−0.8 TBP-RL
Eb /N0 (dB)

TBP-LDPC [2]
−1
MET-LDPC [1]
−1.2

−1.4

−1.6
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
R
Fig. 1: Asymptotic thresholds of the optimized TBP-RL LDPC code distribution at different rates.

100
β = 0.95
Frame Error Rate

β = 0.94
10−1
β = 0.93
P/S R = 0.1
10−2
P/S R = 0.05
P/S R = 0.02
10−3
−21 −19 −17 −15 −13 −11 −9
Es /N0 (dB)

Fig. 2: Frame error rate simulation of the optimized TBP-RL LDPC code at different fixed reconciliation efficiencies β . For
comparison, the performances of randomly punctured and shortened (P/S) fixed rate codes of rates 0.02, 0.05 and 0.1 for
β = 0.95 are plotted.
LDPC code for all rates within the range of 0.01 to 0.2. Therefore, with a single PCM, the TBP-RL LDPC code
can operate across a wide range of rates without sacrificing performance.
To assess the finite-length performance of the designed code, the optimized protograph is lifted to generate
a PCM with a block length of N = 2 · 106 . Figure 2 presents the Monte-Carlo simulation results of the lifted
TBP-RL LDPC code for the binary input additive white Gaussian noise channel at various fixed reconciliation
efficiencies β . The FER performance of the optimized TBP-RL LDPC code remains stable across different rates R
and reconciliation efficiencies β , demonstrating the effectiveness of the proposed code. For comparison, random
puncturing and shortening of fixed-rate codes with base rates of 0.02, 0.05, and 0.1 for β = 0.95 are also simulated.
The results indicate that puncturing and shortening fixed-rate codes yield inferior performance compared to the
TBP-RL LDPC code. Furthermore, more than 3 discrete codes with puncturing and shortening are required to
cover the entire rate range of [0.01, 0.2].
4. Conclusion
This paper presents a novel method for designing raptor-like LDPC codes, referred to as TBP-RL LDPC codes,
using type-based protographs. By optimizing the highest rate code and incrementally extending the IRC proto-
graph using the pre-defined types, a TBP-RL LDPC code can be easily optimized to operate across a wide range
of rates without sacrificing performance. To facilitate the adoption of the designed code, we have made it available
online for use in the IR of CV-QKD systems [7].
Acknowledgement: This work was funded by the German Federal Ministry of Education and Research (BMBF) under grant agreement
16KISQ056 (DE-QOR).
References
1. H. Mani, “Error reconciliation protocols for continuous-variable quantum key distribution,” Ph.D. thesis, Technical University of
Denmark (2021).
2. K. Gümüş and L. Schmalen, “Low rate protograph-based LDPC codes for continuous variable quantum key distribution,” in Proc.
International Symposium on Wireless Communication Systems (ISWCS), (2021).
3. X. Wang, Y. Zhang, S. Yu, B. Xu, Z. Li, and H. Guo, “Efficient rate-adaptive reconciliation for continuous-variable quantum key
distribution,” Quantum Inf. Comput. 17, 1123–1134 (2017).
4. C. Zhou, X. Wang, Z. Zhang, S. Yu, Z. Chen, and H. Guo, “Rate compatible reconciliation for continuous-variable quantum key
distribution using Raptor-like LDPC codes,” Sci. China Physics, Mech. & Astron. 64, 260311 (2021).
5. T.-Y. Chen, K. Vakilinia, D. Divsalar, and R. D. Wesel, “Protograph-based raptor-like LDPC codes,” IEEE Trans. on Commun. 63,
1522–1532 (2015).
6. ETSI 3GPP, “5G NR multiplexing and channel coding,” Tech. Rep. TS 38.212, ETSI (2021).
7. E. E. Cil and L. Schmalen, “Rate-Adaptive Protograph-Based Raptor-Like LDPC Code for Continuous-Variable Quantum Key Dis-
tribution,” https://github.com/erdemeray/TBPRL-Codes (2024). [Online].