Quantization of high dimensional Gaussian vector using permutation modulation with application to information reconciliation in continuous variable QKD
This paper is focused on the problem of Information Reconciliation (IR) for continuous variable Quantum Key Distribution (QKD). The main problem is quantization and assignment of labels to the samples of the Gaussian variables observed at Alice and Bob. Trouble is that most of the samples, assuming that the Gaussian variable is zero mean which is de-facto the case, tend to have small magnitudes and are easily disturbed by noise. Transmission over longer and longer distances increases the losses corresponding to a lower effective Signal-to-Noise Ratio (SNR) exasperating the problem. Quantization over higher dimensions is advantageous since it allows for fractional bit per sample accuracy which may be needed at very low SNR conditions whereby the achievable secret key rate is significantly less than one bit per sample. In this paper, we propose to use Permutation Modulation (PM) for quantization of Gaussian vectors potentially containing thousands of samples. PM is applied to the magnitudes of the Gaussian samples and we explore the dependence of the sign error probability on the magnitude of the samples. At very low SNR, we may transmit the entire label of the PM code from Bob to Alice in Reverse Reconciliation (RR) over public channel. The side information extracted from this label can then be used by Alice to characterize the sign error probability of her individual samples. Forward Error Correction (FEC) coding can be used by Bob on each subset of samples with similar sign error probability to aid Alice in error correction. This can be done for different subsets of samples with similar sign error probabilities leading to an Unequal Error Protection (UEP) coding paradigm.
This paper is focused on the problem of Information Reconciliation (IR) for continuous variable Quantum Key Distribution (QKD). The main problem is quantization and assignment of labels to the samples of the Gaussian variables observed at Alice and Bob. Trouble is that most of the samples, assuming that the Gaussian variable is zero mean which is de-facto the case, tend to have small magnitudes and are easily disturbed by noise. Transmission over longer and longer distances increases the losses corresponding to a lower effective Signal-to-Noise Ratio (SNR) exasperating the problem. Quantization over higher dimensions is advantageous since it allows for fractional bit per sample accuracy which may be needed at very low SNR conditions whereby the achievable secret key rate is significantly less than one bit per sample. In this paper, we propose to use Permutation Modulation (PM) for quantization of Gaussian vectors potentially containing thousands of samples. PM is applied to the magnitudes of the Gaussian samples and we explore the dependence of the sign error probability on the magnitude of the samples. At very low SNR, we may transmit the entire label of the PM code from Bob to Alice in Reverse Reconciliation (RR) over public channel. The side information extracted from this label can then be used by Alice to characterize the sign error probability of her individual samples. Forward Error Correction (FEC) coding can be used by Bob on each subset of samples with similar sign error probability to aid Alice in error correction. This can be done for different subsets of samples with similar sign error probabilities leading to an Unequal Error Protection (UEP) coding paradigm.
This paper is focused on the problem of Information Reconciliation (IR) for continuous variable Quantum Key Distribution (QKD). The main problem is quantization and assignment of labels to the samples of the Gaussian variables observed at Alice and Bob. Trouble is that most of the samples, assuming that the Gaussian variable is zero mean which is de-facto the case, tend to have small magnitudes and are easily disturbed by noise. Transmission over longer and longer distances increases the losses corresponding to a lower effective Signal to Noise Ratio (SNR) exasperating the problem. Here we propose to use Permutation Modulation (PM) as a means of quantization of Gaussian vectors at Alice and Bob over a d-dimensional space with d ≫ 1. The goal is to achieve the necessary coding efficiency to extend the achievable range of continuous variable QKD by quantizing over larger and larger dimensions. Fractional bit rate per sample is easily achieved using PM at very reasonable computational cost. Ordered statistics is used extensively throughout the development from generation of the seed vector in PM to analysis of error rates associated with the signs of the Gaussian samples at Alice and Bob as a function of the magnitude of the observed samples at Bob.
This paper is focused on the problem of Information Reconciliation (IR) for continuous variable Quantum Key Distribution (QKD). The main problem is quantization and assignment of labels to the samples of the Gaussian variables observed at Alice and Bob. Trouble is that most of the samples, assuming that the Gaussian variable is zero mean which is de-facto the case, tend to have small magnitudes and are easily disturbed by noise. Transmission over longer and longer distances increase the losses corresponding to a lower effective Signal to Noise Ratio (SNR) exasperating the problem. In previous work [1], [2] we approach the IR problem in a fundamentally different way than how it is addressed in published literature. At very low SNR, one can focus exclusively on the sign of the Gaussian samples at Alice and Bob and perform Reverse Reconciliation (RR) on sign of the samples given their quantized magnitudes as side information Alice and Bob share in the open. In this paper we demonstrate that the weighted sum capacity of the Binary Symmetric Channels (BSCs) derived from magnitude dependent sign error probabilities of the Gaussian samples shared between Alice and Bob approaches the mutual information between Alice and Bob at very low SNR regimes. Thus, when SNR is very low, one could just as well ignore any common randomness that may be obtained from magnitude of the Gaussian samples and focus exclusively on their sign. The caveat is, to get close to this mutual information, Unequal Error Protection (UEP) must be employed based on magnitude of the observed samples at Bob in RR.
The main problem for information reconciliation in continuous variable Quantum Key Distribution (QKD) at low Signal to Noise Ratio (SNR) is quantization and assignment of labels to the samples of the Gaussian Random Variables (RVs) observed at Alice and Bob. Trouble is that most of the samples, assuming that the Gaussian variable is zero mean which is de-facto the case, tend to have small magnitudes and are easily disturbed by noise. Transmission over longer and longer distances increases the losses corresponding to a lower effective SNR exasperating the problem. This paper looks at the quantization problem of the Gaussian samples at very low SNR regime from an information theoretic point of view. We look at the problem of two bit per sample quantization of the Gaussian RVs at Alice and Bob and derive expressions for the mutual information between the bit strings as a result of this quantization. The quantization threshold for the Most Significant Bit (MSB) should be chosen based on the maximization of the mutual information between the quantized bit strings. Furthermore, while the LSB string at Alice and Bob are balanced in a sense that their entropy is close to maximum, this is not the case for the second most significant bit even under optimal threshold. We show that with two bit quantization at SNR of -3 dB we achieve 75.8% of maximal achievable mutual information between Alice and Bob, hence, as the number of quantization bits increases beyond 2-bits, the number of additional useful bits that can be extracted for secret key generation decreases rapidly. Furthermore, the error rates between the bit strings at Alice and Bob at the same significant bit level are rather high demanding very powerful error correcting codes. While our calculations and simulation shows that the mutual information between the LSB at Alice and Bob is 0.1044 bits, that at the MSB level is only 0.035 bits. Hence, it is only by looking at the bits jointly that we are able to achieve a mutual information of 0.2217 bits which is 75.8% of maximum achievable. The implication is that only by coding both MSB and LSB jointly can we hope to get close to this 75.8% limit. Hence, non-binary codes are essential to achieve acceptable performance.
Quantum Key Distribution (QKD) is a communication method which exchanges secret keys using cryptographic protocols involving elements from quantum science. Continuous Variable (CV) QKD is a method to implement key exchange using sampling of Gaussian signals. Reconciliation in CV-QKD is fundamentally realized via coding of the Alice or Bob binary labels of the Gaussian samples using either one-way or interactive communications between the parties, Alice and Bob, over a public authenticated channel. We assume that the communication is performed through an Optical Wireless (OW) or Free Space Optics (FSO) channel. In that case the received signal suffers from stochastic fading due to pointing jitter or atmospheric turbulence. As a result of the channel fading and noise Alice and Bob Gaussian samples will not match. Information Reconciliation (IR) is the phase of the CV-QKD protocol that makes sure that Alice and Bob agree on a common and identical labeling of their samples, i.e. agree on a common stream of bits that we denote as “reconciled key.” The information reconciliation in CV-QKD could be done by one-way or two-way channel coding to correct the actual labels so that they do match. To do so Bob generates a sequence of parity bits using a systematic code. These bits are sent to Alice over an authenticated public channel. Alice then uses her own sequence of labels she obtains after quantization with the redundancy provided by Bob to recover Bob’s binary sequence. In this work we analyze the problem of information reconciliation for continuous variable quantum key distribution over a free space optics channel.
Continuous-variable quantum key distribution (CV-QKD) requires highly efficient reconciliation techniques to operate at low signal-to-noise ratios and long distances. Multidimensional reconciliation addresses this challenge by transforming the physical Gaussian quantum channel into a virtual binary-input additive white Gaussian noise (BIAWGN) channel, enabling the use of modern errorcorrecting codes. In this work, we review the principles of multidimensional reconciliation, with a particular focus on high-dimensional constructions beyond the algebraic dimensions 1, 2, 4, 8. We describe the construction of the virtual channel, discuss practical coding schemes for reverse reconciliation, and analyse their integration with linear error-correcting codes. We also present an opensource simulation framework, HDirac, implementing multidimensional reconciliation for arbitrary dimensions, and use it to evaluate state-of-the-art LDPC codes. The results highlight key trade-offs between dimension, reconciliation efficiency, and frame error rate, providing practical guidance for CV-QKD system design.
Continuous variable quantum key distribution (CV-QKD) allows the distribution of symmetric keys in a secure manner. CV-QKD systems can extract keys at its maximum rate when using Gaussian modulation (GM). Nonetheless, GM demands high-capacity random number sources and tends to be very hard to approach in practice. To circumvent these disadvantages, higher-order probabilistic -shaped discrete modulation (DM) can be used. State-of-the-art works compute the key rates of DM-CV-QKD systems considering a fixed value for the reconciliation efficiency and do not take into account the frame error rate (FER) of the system, thus over or under estimating the key rates. In this work, we study the security bounds of CV-QKD systems considering probabilistic shaped DM formats with 256-symbols in the finite-size regime. This accounting for the true value of the reconciliation efficiency, and the FER of the information reconciliation step. Both conventional and hexagonal 256-quadrature amplitude modulation (QAM) constellations yield higher key rates than 256-amplitude and phase shift keying (APSK) constellations, with 256-QAM constellations being indistinguishable in performance with GM for high transmission distances. Minimum signal-to-noise ratio (SNR) values were fixed from a FER analysis through a CV-QKD simulation allowing for key extraction considering different FER levels. Lower FER values are associated with higher SNRs in the system and thus lower achievable transmission distances. Nonetheless, the FER maximizing the extraction key rate is not null. Our results show that the extraction key rate is maximized by SNR adjustment which should have in account both the reconciliation efficiency and the FER.