Short Codes Research Articles

Low-Latency Nested Decoding for Short Generalized Integrated Interleaved BCH Codes

Generalized integrated interleaved (GII) codes nest short BCH sub-codewords to form more powerful BCH codewords. They can potentially achieve hyper-speed decoding with excellent error-correction capability. In particular, short GII-BCH codes are among the best candidates for the new fast storage class memories (SCMs). Miscorrections severely degrade the performance of short GII-BCH codes. Although they were effectively mitigated in previous designs, the involved repeated Chien search and higher-order syndrome computation cause long latency. This brief proposes efficient and low-latency nested decoding schemes for short GII-BCH codes. A strategy is developed to select sub-words for further nested decoding to mitigate miscorrections by keeping track of the error locator polynomials, instead of waiting for the lengthy Chien search. Formulas are also derived to estimate the effects on the error-correcting performance. Besides, a low-complexity linear feedback shift register (LFSR) architecture is developed to accelerate the higher-order nested syndrome computation. For an example GII-BCH code targeting at SCMs, the proposed design reduces the worst-case nested decoding latency by 26% with 8.5% area overhead and negligible performance loss compared to prior methods.

Watcher: Cloud-Based Coding Activity Tracker for Fair Evaluation of Programming Assignments.

Online learning has made it possible to attend programming classes regardless of the constraint that all students should be gathered in a classroom. However, it has also made it easier for students to cheat on assignments. Therefore, we need a system to deal with cheating on assignments. This study presents a Watcher system, an automated cloud-based software platform for impartial and convenient online programming hands-on education. The primary features of Watcher are as follows. First, Watcher offers a web-based integrated development environment (Web-IDE) that allows students to start programming immediately without the need for additional installation and configuration. Second, Watcher collects and monitors the coding activity of students automatically in real-time. As Watcher provides the history of the coding activity to instructors in log files, the instructors can investigate suspicious coding activities such as plagiarism, even for a short source code. Third, Watcher provides facilities to remotely manage and evaluate students’ hands-on programming assignments. We evaluated Watcher in a Unix system programming class for 96 students. The results showed that Watcher improves the quality of the coding experience for students through Web-IDE, and it offers instructors valuable data that can be used to analyze the various coding activities of individual students.

Beyond 100 Gbit/s Pipeline Decoders for Spatially Coupled LDPC Codes

Low-density parity-check (LDPC) codes are a well-established class of forward error correction codes that provide excellent error correction performance for large code block sizes. However, for throughputs toward 1 Tbit/s, as expected for B5G systems, state-of-the-art LDPC soft decoders are restricted to short code block sizes of several hundred to thousands of bits due to routing congestion challenges, limiting the overall communications performance of the transmission system. Spatially coupled LDPC (SC-LDPC) codes and respective sliding window decoding methods show the potential to overcome these block size restrictions. However, in contrast to conventional LDPC codes little literature exists on the efficient hardware implementation of respective high-throughput decoders. In this work, we present the first in-depth investigation on the implementation of SC-LDPC decoders for throughputs beyond 100 Gbit/s. For an N=51328, R=0.8 terminated SC-LDPC code with sub-block size c=640 and coupling width m_s=1, we explore various design trade-offs, including row- and column-wise decoding, non-overlapping and overlapping window scheduling, and processor pipelining. To the best of our knowledge, this is the first description of a column-wise SC-LDPC decoding architecture in the literature. We complement the algorithmic investigation with the virtual silicon implementation of all presented decoders in a 22nm FD-SOI technology.

Short Non-Malleable Codes from Related-Key Secure Block Ciphers, Revisited

We construct non-malleable codes in the split-state model with codeword length m + 3λ or m + 5λ, where m is the message size and λ is the security parameter, depending on how conservative one is. Our scheme is very simple and involves a single call to a block cipher meeting a new security notion which we dub entropic fixed-related-key security, which essentially means that the block cipher behaves like a pseudorandom permutation when queried upon inputs sampled from a distribution with sufficient min-entropy, even under related-key attacks with respect to an arbitrary but fixed key relation. Importantly, indistinguishability only holds with respect to the original secret key (and not with respect to the tampered secret key).In a previous work, Fehr, Karpman, and Mennink (ToSC 2018) used a related assumption (where the block cipher inputs can be chosen by the adversary, and where indistinguishability holds even with respect to the tampered key) to construct a nonmalleable code in the split-state model with codeword length m + 2λ. Unfortunately, no block cipher (even an ideal one) satisfies their assumption when the tampering function is allowed to be cipher-dependent. In contrast, we are able to show that entropic fixed-related-key security holds in the ideal cipher model with respect to a large class of cipher-dependent tampering attacks (including those which break the assumption of Fehr, Karpman, and Mennink).

Overcoming Data Availability Attacks in Blockchain Systems: Short Code-Length LDPC Code Design for Coded Merkle Tree

Light nodes are clients in blockchain systems that only store a small portion of the blockchain ledger. In certain blockchains, light nodes are vulnerable to a data availability (DA) attack where a malicious node makes the light nodes accept an invalid block by hiding the invalid portion of the block from the nodes in the system. Recently, a technique based on LDPC codes called Coded Merkle Tree was proposed by Yu et al. that enables light nodes to detect a DA attack by randomly requesting/sampling portions of the block from the malicious node. However, light nodes fail to detect a DA attack with high probability if a malicious node hides a small stopping set of the LDPC code. In this paper, we demonstrate that a suitable co-design of specialized LDPC codes and the light node sampling strategy leads to a high probability of detection of DA attacks. We consider different adversary models based on their computational capabilities of finding stopping sets. For the different adversary models, we provide new specialized LDPC code constructions and coupled light node sampling strategies and demonstrate that they lead to a higher probability of detection of DA attacks compared to approaches proposed in earlier literature.

Rapid Person Re-Identification via Sub-space Consistency Regularization

Person Re-Identification (ReID) matches pedestrian across disjoint cameras. Existing ReID methods adopting real-value feature descriptors have achieved high accuracy, but they are low in efficiency due to the slow Euclidean distance computation as well as complex quick-sort algorithms. Recently, some works propose to yield binary encoded person descriptors which instead only require fast Hamming distance computation and simple counting-sort algorithms. However, the performances of such binary encoded descriptors, especially with short code (e.g, 32 and 64 bits), are hardly satisfactory given the sparse binary space. To strike a balance between the model accuracy and efficiency, we propose a novel Sub-space Consistency Regularization (SCR) algorithm that can speed up the ReID procedure by 0.25 times than real-value features under same dimensions whilst maintain a competitive accuracy, especially under short codes. SCR transforms real-value features vector (e.g, 2048 float32) with short binary codes (e.g, 64 bits) by first dividing real-value features vector into M sub-spaces, each with C clustered centroids. Thus the distance between two samples can be expressed as the summation of respective distance to the centroids, which can be sped up by offline calculation and maintained via a look-up-table. On the other side, these real-value centroids help to achieve significantly higher accuracy than using binary code. Lastly, we convert the distance look-up-table to be integer and apply the counting-sort algorithm to speed up the ranking stage. We also propose a novel consistency regularization with an iterative framework. Experimental results on Market-1501 and DukeMTMC-reID show promising and exciting results. Under short code, our proposed SCR enjoys Real-value-level accuracy and Hashing-level speed.

CRC-Aided Adaptive BP Decoding of PAC Codes.

Although long polar codes with successive cancellation decoding can asymptotically achieve channel capacity, the performance of short blocklength polar codes is far from optimal. Recently, Arıkan proposed employing a convolutional pre-transformation before the polarization network, called polarization-adjusted convolutional (PAC) codes. In this paper, we focus on improving the performance of short PAC codes concatenated with a cyclic redundancy check (CRC) outer code, CRC-PAC codes, since error detection capability is essential in practical applications, such as the polar coding scheme for the control channel. We propose an enhanced adaptive belief propagation (ABP) decoding algorithm with the assistance of CRC bits for PAC codes. We also derive joint parity-check matrices of CRC-PAC codes suitable for iterative BP decoding. The proposed CRC-aided ABP (CA-ABP) decoding can effectively improve error performance when partial CRC bits are used in the decoding. Meanwhile, the error detection ability can still be guaranteed by the remaining CRC bits and adaptive decoding parameters. Moreover, compared with the conventional CRC-aided list (CA-List) decoding, our proposed scheme can significantly reduce computational complexity, to achieve a better trade-off between the performance and complexity for short PAC codes.

Abstract P1006: Cross-lineage Potential Of Ascl1 Uncovered By Comparing Diverse Reprogramming Regulatomes

Direct reprogramming has revolutionized the fields of stem cell biology and regenerative medicine, making it possible to interconvert somatic cell fates without intermediary pluripotency. Although studies have identified different transcription factor (TF) cocktails that can reprogram fibroblasts into different cell types, the common mechanisms governing how reprogramming cells undergo transcriptome and epigenome remodeling (i.e., regulatome remodeling) have not been investigated. Here, by characterizing early changes in the regulatome of three different types of direct reprogramming - induced neurons, induced hepatocytes, and induced cardiomyocytes - we identify lineage-specific features as well as common regulatory transcription factors that act in a context-dependent manner for each type of reprogramming. Of particular interest, we discover that the neuronal factor Ascl1 possesses cross-lineage potential; together with Mef2c, it navigates reprogramming fibroblasts through a targeted and potent iCM trajectory. Single-cell multi-omics reveal that A+M reprogramming terminates in a more mature iCM phenotype than MGT. Finally, through ChIP-seq and RNA-seq, we find that Mef2c drives the shift in Ascl1 binding away from neuronal genes towards cardiac genes, guiding their co-operative epigenetic and transcription activities. Given the relatively short DNA coding sequence of Ascl1 (696bp) and Mef2c (1299bp), our novel A+M reprogramming cocktail has the potential to be delivered within a single AAV vector. Such a strategy is not feasible with classic cocktail MGT (4182bp). Together, our study demonstrates the existence of common regulators of different direct reprogramming processes, argues against the premise that TFs possess only lineage-specific capabilities for altering cell fate - the basic premise used to develop direct reprogramming approaches and discover a novel cardiac reprogramming cocktail showing great potential for future clinical application.

Polyphase codes for multiplexed acoustic signalling and sensing on pipes

Transmission of acoustic signals between distributed sensor nodes may be useful for status monitoring of elongated structures such as pipelines. In principle, coded signals can be used in an asynchronous multiplexed system, provided the signals are distinguishable. However, multimode effects complicate signal propagation, so any such codes should be short. A search for polyphase code families with properties suitable for acoustic code division multiple access is presented. Algorithms for reduction of search space to allow use of a laptop for code discovery are described. Short codes of base 6 are shown to outperform codes of bases 2, 3, 4 and sets suitable for systems with 2 and 3 users are identified. The codes have similar properties to Barker codes but larger sidelobes. Their use is demonstrated by simulation and experiment at kHz frequencies using an air-filled copper pipe, an electromagnetic acoustic transducer (EMAT) and a microphone designed to excite and detect the mode. Low loss propagation over 25 m is achieved with a 20 kHz carrier. Excellent agreement between experiment and theory is demonstrated, with performance limited by transducer bandwidth.

Analysis and Design of Short Analog Fountain Codes for the Multiple Access Channel

In this letter, we design and analyse analog fountain codes (AFC) for the multiple access channel (MAC). First, we present several encoding schemes, including the superposition coding and joint encoding, and decoding schemes, including successive interference cancellation and joint decoding, for multiple access AFC (MA-AFC). We propose different combinations of encoding and decoding schemes for MA-AFC and evaluate their performance in terms of block error rate (BLER) to determine which combination has the best performance. We propose a density evolution (DE) based framework to track the evolution of messages exchanged between check nodes and variable nodes in the iterative joint belief propagation (BP) decoder of MA-AFC. With the proposed DE framework, we formulate optimisation problems to find the optimal AFC code parameters, specifically the weight set, which minimises the bit error rate at a given signal-to-noise ratio (SNR). Simulation results show that the AFC code with the optimised parameters outperforms existing AFC code designs in the multiple access scenario.

Frequency Sweep Keying CDMA for Reducing Ultrasonic Crosstalk

Various sensors are embedded in automobiles to implement intelligent safety technologies such as autonomous driving and front–rear collision avoidance technology. In particular, ultrasonic sensors have been used in the past because they have an accuracy of centimeters to sub-centimeters in air despite their low cost and low hardware complexity. Recently, the crosstalk problem between ultrasonic sensors has been raised because the number of ultrasonic sensors in the unit space has increased as the number of vehicles increases. Various studies have been conducted to solve the crosstalk, but a demodulation error occurs when signals overlap. Therefore, in this paper, we propose a method that is robust to ultrasonic signal overlap, is robust even at shorter code length, and has reduced time of flight (TOF) error compared to the existing method by applying frequency sweep keying modulation based on code division multiple access (CDMA). As a result of the experiment, the code was detected accurately regardless of the overlap ratio of the two signals, and it was robust even in situations where the power of the two signals was different. In addition, it shows an accurate TOF estimation even if the ID code length is shorter than the existing on–off-keying, frequency shift keying, and phase shift keying methods.

Girth aware normalized min sum decoding algorithm for shorter length low density parity check codes

Recently, short block codes are in great demand due to the emergent applications requiring the transmission of a short data unit and can guarantee speedy communication, with a minimum of latency and complexity which are among the technical challenges in today’s wireless services and systems. In the context of channel coding using low density parity check (LDPC) codes, the shorter length LDPC block codes are more likely to have short cycles with lengths of 4 and 6. The effect of the cycle with the minimum size is that this one prevents the propagation of the information in the Tanner graph during the iterative process. Therefore, the message decoded by short block code is assumed to be of poor quality due to short cycles. In this work, we present a study of the evolution of the messages on check nodes during the iterative decoding process when using the LDPC decoding algorithm normalized min sum (NMS), to see the destructive effect of short cycles and justify the effectiveness of the girth aware normalized min sum (GA-NMS) decoding LDPC codes algorithm in terms of correction of the errors, particularly for the codes with short cycles 4 and 6. In addition to this, the GA-NMS algorithm is evaluated in terms of bit error rate performance and convergence behavior, using wireless regional area networks (WRAN) LDPC code, which is considered as a short block code.

CRC-Aided List Decoding of Convolutional Codes in the Short Blocklength Regime

We consider the concatenation of a convolutional code (CC) with an optimized cyclic redundancy check (CRC) code as a promising paradigm for good short blocklength codes. The resulting CRC-aided convolutional code naturally permits the use of serial list Viterbi decoding (SLVD) to achieve maximum-likelihood decoding. The convolutional encoder of interest is of rate-$1/\omega$ and the convolutional code is either zero-terminated (ZT) or tail-biting (TB). The resulting CRC-aided convolutional code is called a CRC-ZTCC or a CRC-TBCC. To design a good CRC-aided convolutional code, we propose the distance-spectrum optimal (DSO) CRC polynomial. A DSO CRC search algorithm for the TBCC is provided. Our analysis reveals that the complexity of SLVD is governed by the expected list rank which converges to $1$ at high SNR. This allows a good performance to be achieved with a small increase in complexity. In this paper, we focus on transmitting $64$ information bits with a rate-$1/2$ convolutional encoder. For a target error probability $10^{-4}$, simulations show that the best CRC-ZTCC approaches the random-coding union (RCU) bound within $0.4$ dB. Several CRC-TBCCs outperform the RCU bound at moderate SNR values.

Randomized Spectral Separation Coefficient for Short Code Acquisition Performance Evaluation

Reliable signal acquisition with low computational complexity is an important design objective for the evolution of global navigation satellite systems (GNSS). Most GNSS signals consist of long pseudorandom noise (PRN) codes whose acquisition is expensive in terms of memory, computation time, and energy. As these resources are particularly scarce in the emerging mass-market user segment, cyclostationary pilot signals with short PRN codes are an attractive option to keep the number of acquisition search bins low. However, reducing the code length degrades the acquisition performance, as multiple access interference (MAI) becomes more pronounced and can lead to an increased false alarm rate. We demonstrate that, quite different from stationary MAI, cyclostationary MAI does not affect each bin of the search space uniformly, and is therefore not easily modeled with the well-known spectral separation coefficient (SSC). We propose a new randomized SSC (SSC-R) based on code/Doppler interference functions, which can be used for simple and accurate acquisition performance evaluation. As an application example, we demonstrate how the SSC-R can be utilized in signal design to minimize the PRN code length under an acquisition performance constraint. We conclude that feasible PRN code lengths for GNSS can be on the order of 300-700.

Keep the Bursts and Ditch the Interleavers

While many communications media, such as wireless and certain classes of wireline channels, typically lead to bursty errors, most decoders are designed assuming memoryless channels. Consequently, communication systems generally rely on interleaving over tens of thousands of bits to match decoder assumptions. Even for short high rate codes, awaiting sufficient data in interleaving and de-interleaving is a significant source of unwanted latency. We construct an extension to the recently proposed Guessing Random Additive Noise Decoding (GRAND) algorithm, which we call GRAND-MO for GRAND Markov Order. By foregoing interleaving and instead making use of the bursty nature of noise, low-latency communication is possible with block error rates outperforming their interleaved counterparts by a substantial margin. We establish that certain well-known binary codes with structured code-word patterns are ill-suited for use in bursty channels, but Random Linear Codes (RLCs) prove robust to correlated noise. We further demonstrate that by operating directly on modulated symbols rather than de-mapped bits, GRAND-MO achieves further performance and complexity gains by exploiting information that is lost in demodulation. As a result, GRAND-MO provides one potential solution for applications that require ultra-reliable low latency communication.

A Simple Method to Calculate Random-Coding Union Bounds for Ultra-Reliable Low-Latency Communications

Ultra-Reliable Low-Latency Communications are based on very short codes whose performances cannot be evaluated correctly by using the Shannon capacity formula holding for asymptotically large code lengths. The use of Random-Coding Union Bounds (RCUB’s) has been suggested in the literature as an alternative for this application. Unfortunately, their calculation is difficult and the Gaussian approximation may lead to erroneous results. The saddlepoint approximation has been proposed as an alternative to overcome the limitations of the Gaussian approximation. Though this technique is valid in many cases, situations exist where the exact calculation provides different results. A simple numerical technique is proposed in this letter to calculate numerically the exact value of the RCUB. Its accuracy is compared to that of the Gaussian and saddlepoint approximations in some cases of interest.

Performance enhanced probabilistically shaped PAM4 signal based on nonlinear differential coding for IM/DD system

In this paper, the probabilistically shaped pulse amplitude modulation (PS-PAM) signal is proposed in intensity modulation and direct detection (IM/DD) system. For the traditional shaping scheme with constant composition distribution matcher (CCDM), there are several problems that high-precision algorithm for long block coding, the loss of information rate for short block coding. In addition, it can increase nonlinear interference. In order to reduce computational complexity and alleviate the influence of nonlinear interference on the system, the nonlinear differential coding (NLDC) scheme is used to realize PS, which can be better integrated with the cost-effective optical access network. This paper evaluates seven shaping spectrums obtained by different NLDC coding coefficients. As the coding coefficient increases, the performance of the system gradually improves. Moreover, this paper proposed the machine learning includes neural network (NN) and improved K-means clustering algorithm at the receiver. The results show that the PS-PAM4 signal is significantly better than the ordinary PAM4 signal in improving the channel bandwidth tolerance and signal transmission rate. Under the same data rate, using NLDC can save 5.8 GHz bandwidth for the system. For the bandwidth-limited system, the PS-PAM4 signal can obtain 19Gbit/s bit rate gain compared with the ordinary PAM4 signal. Compared with the typical shaping scheme using CCDM, this scheme can achieve 2.2 GHz band tolerance gain and 5Gbit/s rate gain in nonlinear fiber channel.

Delving into the representation learning of deep hashing

Searching for the nearest neighbor is a fundamental problem in the computer vision field, and deep hashing has become one of the most representative and widely used methods, which learns to generate compact binary codes for visual data. In this paper, we first delve into the representation learning of deep hashing and surprisingly find that deep hashing could be a double-edged sword, i.e., deep hashing can accelerate the query speed and decrease the storage cost in the nearest neighbor search progress, but it greatly sacrifices the discriminability of deep representations especially with extremely short target code lengths. To solve this problem, we propose a two-step deep hashing learning framework. The first step focuses on learning deep discriminative representations with metric learning. Subsequently, the learning framework concentrates on simultaneously learning compact binary codes and preserving representations learned in the former step from being sacrificed. Extensive experiments on two general image datasets and four challenging image datasets validate the effectiveness of our proposed learning framework. Moreover, the side effect of deep hashing is successfully mitigated with our learning framework.

Perceptual Hash of Neural Networks

In recent years, advances in deep learning have boosted the practical development, distribution and implementation of deep neural networks (DNNs). The concept of symmetry is often adopted in a deep neural network to construct an efficient network structure tailored for a specific task, such as the classic encoder-decoder structure. Massive DNN models are diverse in category, quantity and open source frameworks for implementation. Therefore, the retrieval of DNN models has become a problem worthy of attention. To this end, we propose a new idea of generating perceptual hashes of DNN models, named HNN-Net (Hash Neural Network), to index similar DNN models by similar hash codes. The proposed HNN-Net is based on neural graph networks consisting of two stages: the graph generator and the graph hashing. In the graph generator stage, the target DNN model is first converted and optimized into a graph. Then, it is assigned with additional information extracted from the execution of the original model. In the graph hashing stage, it learns to construct a compact binary hash code. The constructed hash function can well preserve the features of both the topology structure and the semantics information of a neural network model. Experimental results demonstrate that the proposed scheme is effective to represent a neural network with a short hash code, and it is generalizable and efficient on different models.

An efficient dual semantic preserving hashing for cross-modal retrieval

Hashing methods have recently received widespread attention due to their flexibility and effectiveness for cross-modal retrieval tasks. However, most existing cross-modal hashing methods have some challenging problems, in particular, effective exploitation of semantic information and learning discriminative hash codes. To address these challenges, we propose an efficient Dual Semantic Preserving Hashing (DSPH) method, which first leverages matrix factorization to obtain low-level latent semantic representations of different modalities and remove redundant information. To enhance the discriminative capability of hash codes, we preserve the high-level pairwise semantics and the learned low-level latent semantics into the unified hash codes. Finally, DSPH adopts discrete optimization strategy to learn the hash codes directly. Experimental results on three benchmark datasets demonstrate that the proposed DSPH method outperforms many state-of-the-art cross-modal hashing methods in terms of retrieval accuracy, especially when dealing with short hash code.