Search

The Internet of Things (IoT) devices, networks, and applications have become an integral part of modern societies. Despite their social, economic, and industrial benefits, these devices and networks are frequently targeted by cybercriminals. Hence, IoT applications and networks demand lightweight, fast and flexible security solutions to overcome these challenges. In this regard, Artificial Intelligence (AI)-based solutions with big data analytics can produce promising results in the field of cybersecurity. This article proposes a lightweight Dense Random Neural Network (DnRaNN) for intrusion detection in the IoT. The proposed scheme is well suited for implementation in resource-constrained IoT networks due to its inherent improved generalization capabilities and distributed nature. The suggested model was evaluated by conducting extensive experiments on a new generation IoT security dataset ToN_IoT. All the experiments were conducted under different hyperparameters and the efficiency of the proposed DnRaNN was evaluated through multiple performance metrics. The findings of the proposed study provide recommendations and insights in binary class and multiclass scenarios. The proposed DnRaNN model attained attack detection accuracy of 99.14% and 99.05% for binary class and multiclass classifications, respectively.

Block orders are increasingly adopted in many electricity markets as they enable power producers to implicitly express their unit commitment characteristics and address price volatility risks. However, previous work on the optimal offering problem of a price-taking producer participating in such markets has only considered the simplest type of block orders (regular block orders-RBO) and has neglected relevant realistic market regulations. This paper addresses these simplifications by proposing a new optimal offering model which also considers the more complex profile block orders (PBO) and linked block orders (LBO). This model is generic in that it can treat the block orders time spans either as fixed parameters (following the modelling approach of previous work) or as endogenous decision variables, which constitutes a completely novel modeling approach. The presented case studies demonstrate that this novel approach yields a tremendously lower number of binary variables and more efficient computational performance, when PBO and LBO are considered. Furthermore, the case studies analyze the physical conditions under which the introduction of different types of block orders increases the examined producers profit.

Robust watermarking is a valuable methodology used in protecting the copyright and securing digital images. In this paper, new fractional-order multi-channel orthogonal exponent moments (MFrEMs) and their invariants to geometric transformations are derived for the first time. We utilized these highly accurate moments to construct a new robust watermarking algorithm for color images. This algorithm consists of three phases. First, the bits of the binary watermark scrambled by using a 1D Sine chaotic map. Second, the fractional-order MFrEMs are calculated from the host color image. Finally, a quantization process is performed, where the scrambled bits of the binary watermark embedded into the host color image. Various experiments were conducted to test the proposed watermarking algorithm and compare it with the existing robust watermarking algorithms for color images. The obtained results ensure the proposed robust watermarking algorithm’s superiority over existing algorithms regarding the visual imperceptibility and robustness against various attacks.

Sparse convolutional neural network (CNN) models reduce the massive compute and memory bandwidth requirements inherently present in dense CNNs without a significant loss in accuracy. Sparse CNNs, however, present their own set of challenges including non-linear data accesses and complex design of CNN processing elements (PEs). Recently proposed accelerators like SCNN, Eyeriss v2, and SparTen, exploit the two-sided sparsity, that is, sparsity in both the input activations and weights to accelerate the CNN inference. These, accelerators, however, suffer from a multitude of problems that limit their applicability, such as inefficient micro-architecture (SCNN, Eyeriss v2), complex PE design (Eyeriss v2), no support for non-unit stride convolutions (SCNN) and FC layers (SparTen, SCNN). To address these issues in contemporary sparse CNN accelerators, we propose Sparse-PE, a multi-threaded, and flexible CNN PE, capable of handling both the dense and sparse CNNs. The Sparse-PE core uses binary mask representation and actively skips computations involving zeros and favors non-zero computations, thereby, drastically increasing the effective throughput and hardware utilization. Unlike previous designs, the Sparse-PE core is generic in nature and not targeted towards a specific accelerator, and thus, can also be used as a standalone sparse dot product compute engine. We evaluate the performance of the core using a custom built cycle accurate simulator. Our simulations show that the Sparse-PE core-based accelerator provides a performance gain of $12\times $ over a recently proposed dense accelerator (NeuroMAX). For sparse accelerators, it provides a performance gain of $4.2\times $ , $2.38\times $ , and $1.98\times $ over SCNN, Eyeriss v2, and SparTen, respectively.

Fire detection has been an issue of interest to researchers due to its significant damage to lives and property within a very short time. One of the recent solutions developed to detect fire is to use Internet-of-Things (IoT) devices equipped with cameras for surveillance. The captured videos of surroundings may be processed by the IoT devices themselves or at the cloud. The latter case is required if the detection algorithm is computationally demanding. However, the use of cloud has a flaw. In fact, using the cloud could pose the threat of having the privacy of a place violated, either through hacking or unauthorized access to the footage of the place where the cloud is installed. In this paper, a fire detection system that preserves the privacy of surroundings, while maintaining a high level of accuracy for fire detection is proposed. The proposed system makes use of the cloud for fire detection; and that is achieved by sending to the cloud features extracted from the video captured by the IoT device, instead of sending the actual footage. Binary video descriptors and Convolutional Neural Network (CNN) have been used to develop the fire detection algorithm. The video descriptors are used to extract features, while the CNN is used for classification. Videos with real fire and non-fire scenes have been used in this development. Results show that the performance of proposed fire detection algorithm can achieve 97.5% classification accuracy, that outperforms the state-of-the art algorithms which make direct use of raw videos. Therefore, the proposed fire detector is as reliable as other available systems, with the advantage of having a privacy-preserving capability. It is also demonstrated that the proposed video descriptors can be implemented for real-time processing using an IoT device, Raspberry Pi 4 platform, with an average processing speed of 100ms per frame, which well satisfies practical needs.

Recently, a little research into image encryption has been used on chaotic economic maps. The current paper suggests a bit-level permutation and a non-invertible chaotic economic map to encrypt an image. Firstly, the secret key generation is linked to the plain image. So, the suggested algorithm may resist both known-plaintext and chosen-plaintext attacks. Then a bit-level permutation is performed for all the binary bits of the plain image’s pixels, using the logistic map (permutation stage). It is used to improve the algorithm’s security. Then pixel diffusion is applied using the 2D non-invertible chaotic economic map and bit-wise XOR operations. It is used to change the pixels’ values and make them highly random. The results of the experiments and the security analyses show that the given image encryption algorithm is efficient with higher security. Some comparisons showed that the proposed algorithm outperformed many recent algorithms. Finally, the proposed algorithm may be able to withstand a variety of attacks.

Real-time three-dimensional reconstruction is increasingly important in many fields. There are mainly two ways to improve real-time performance: increasing hardware rate and reducing captured frames count. For increasing hardware rate, the binary defocusing technique can effectively break the limitation of speed. For reducing captured frames count, the fusion technique is typically introduced. However, the simultaneous work of binary defocusing technique and fusion technique is hard to achieve. To this end, we establish a novel system framework with a camera and dual projectors. The dual projectors can control defocusing level independently, i.e. enhance the flexibility of projector. Based on this, our paper exhibits two decoding phase methods: the strict-alignment method and the loose-alignment method. The strict-alignment method requires the projection chips to be strictly aligned, and a virtual example demonstrated effectiveness with three captured frames. In order to further improve practicality, the loose-alignment method, which relaxes the restriction, is proposed. Finally, experiments verify the performance of our proposed method.

We report 108 Gb/s Duo-Binary PAM-8 (DB-PAM-8) transmission using direct-detection aided by Volterra equalizer. The bit-error-rate (BER) performance of the DB-PAM-8 signal has been evaluated through error counting by varying the bit rate, received optical power, and reach. The experimental results suggest that up to 108, 102, and 84 Gb/s signals can be propagated in back-to-back (B2B), and over 10, and 20 km of standard single-mode fiber (SMF), respectively, below the low-density parity-check forward error correction (LDPC-FEC) threshold. Furthermore, we derive a probabilistic model of the BER for multilevel DB-PAM-M signaling that is verified using our measurements. The model of this important performance metric turned out to be accurate enough, especially at low BER values, where its use is more convenient.

Seagull Optimization Algorithm (SOA) is a metaheuristic algorithm that mimics the migrating and hunting behaviour of seagulls. SOA is able to solve continuous real-life problems, but not to discrete problems. The eight different binary versions of SOA are proposed in this paper. The proposed algorithm uses four transfer functions, S-shaped and V-shaped, which are used to map the continuous search space into discrete search space. Twenty-five benchmark functions are used to validate the performance of the proposed algorithm. The statistical significance of the proposed algorithm is also analysed. Experimental results divulge that the proposed algorithm outperforms the competitive algorithms. The proposed algorithm is also applied on data mining. The results demonstrate the superiority of binary seagull optimization algorithm in data mining application.

Bit permutations are efficient linear functions often used for lightweight cipher designs. However, they have low diffusion effects, compared to word-oriented binary and maximum distance separable (MDS) matrices. Thus, the security of bit permutation-based ciphers is significantly affected by differential and linear branch numbers (DBN and LBN) of nonlinear functions. In this paper, we introduce a widely applicable method for constructing S-boxes with high DBN and LBN. Our method exploits constructions of S-boxes from smaller S-boxes and it derives/proves the required conditions for smaller S-boxes so that the DBN and LBN of the constructed S-boxes are at least 3. These conditions enable us to significantly reduce the search space required to create such S-boxes. Using the unbalanced-Bridge and unbalanced-MISTY structures, we develop a variety of new lightweight S-boxes that provide not only both DBN and LBN of at least 3 but also efficient bitsliced implementations including at most 11 nonlinear bitwise operations. The new S-boxes are the first that exhibit these characteristics.

The differences in performance among binary-integer encodings in an Ising machine, which can solve combinatorial optimization problems, are investigated. Many combinatorial optimization problems can be mapped to find the lowest-energy (ground) state of an Ising model or its equivalent model, the Quadratic Unconstrained Binary Optimization (QUBO). Since the Ising model and QUBO consist of binary variables, they often express integers as binary when using Ising machines. A typical example is the combinatorial optimization problem under inequality constraints. Here, the quadratic knapsack problem is adopted as a prototypical problem with an inequality constraint. It is solved using typical binary-integer encodings: one-hot encoding, binary encoding, and unary encoding. Unary encoding shows the best performance for large-sized problems.

Binary sequences are widely used in many practical fields, such as radar applications, telecommunications and cryptography. Finding low autocorrelation binary sequences with good peak side-lobe level (PSL) values is a difficult optimization problem. In this paper we present an improved heuristic algorithm for searching low autocorrelation PSL sequences. A heuristic algorithm can find a sequence with a PSL value, which is not necessarily optimal, but is usually near optimal, and the algorithm finds it in a reasonable amount of time. In the experimental work we applied our algorithm to find binary sequences with low PSL values, and made a comparison with the state-of-the-art algorithms from literature. With our algorithm many sequences with the currently best-known PSL values have been improved. We found new sequences with better, i.e., lower, PSL values.

Locally repairable codes (LRCs) with multiple recovering sets are highly demanded in distributed storage systems. In this letter, we generalize the construction of WZL code proposed by Wang et al. and give a construction of optimal binary LRCs with multiple disjoint recovering sets which can reach the upper bound on the code rate given by Kadhe et al.. Then we further generalize our idea to obtain a construction of binary LRCs with intersect recovering sets. The code rate is much higher than that of WZL code and is very closed to the construction of Kruglik et al. Moreover, two special cases of this construction can reach the upper bound on the minimum distance.

Deep learning (DL) algorithms are crucial for predicting various diseases because they can analyze a large amount of healthcare data within a short prediction time. One of these diseases is cancer, which causes one out of six deaths worldwide. Many researchers have adopted predictive frameworks such as machine learning and DL to predict cancer prognosis, in addition to the probability of its recurrence, progression, and the patients’ survival estimation. Currently, all stakeholders are interested in the accuracy of cancer prognosis prediction. This study selected a framework within high accuracy and short prediction time from three DL frameworks for improving the performance of cancer prognosis prediction. This prediction requires a quick and high-accuracy optimizer, so we propose a binary version of the continuous AC-parametric whale optimization algorithm. This version is built on S-shaped transfer functions to identify the minimal optimal subset of features and maximize the classification accuracy. These frameworks proposed have the following forms: the first is a Feed-Forward Neural Network (FFNN) in which the input is the optimal set of feature selection. The second is an optimized parameter FFNN. The third is composed of a feature selection layer in which the best subset of selected features is for use as inputs in the optimized FFNN. We compared these frameworks using a comparative study. Our results show that, under all conditions, the third framework is superior to the others with an average accuracy of 100%, whereas the first and second frameworks achieved 94.97% and 93.12% accuracy, respectively.

This paper proposes a novel nanoelectromechanical-switch-based binary content-addressable memory (NEMBCAM) for the first time. A nanoelectromechanical (NEM) memory switch serves as both a nonvolatile memory and a switchable current path in an NEMBCAM unit cell. Owing to monolithic three-dimensional integration, NEMBCAM achieves a smaller unit cell area in comparison with conventional complementary metal–oxide–semiconductor-only binary content-addressable memory. Small unit cell area results in reduced match line (ML) capacitance, which lowers search energy consumption. Furthermore, a shorter search delay is achieved owing to the near-perfect on/off characteristic of the NEM memory switch.

In this study, the branch-length similarity entropy profile is estimated by mapping the time-series signal to the circumference of the time circle, and the self-similarity is defined based on the profile. To explore the self-similarity property, the effect of the distance between two signals, “0” and “1”, on the entropy value for signal “1” is investigated. Furthermore, two application problems are addressed: quantification of the mixing state of fragments and clusters, and characterization of the behavioral trajectory of an organism. The results indicate that use of the self-similarity property solves both the problems. Additionally, the problems that must be addressed to broaden the applicability of self-similarity are discussed.

As the third generation of neural networks, spiking neural network (SNN) motivated by neurophysiology enjoys considerable advances due to integrating different information, such as time and space. The frequency-domain provides a powerful capability of modeling and training convolutional neural networks (CNNs). However, SNN with binary input and output will lose much information and slightly inferior to deep neural networks (DNN). We consider how to make the most of information to protect input. Binary input and output are different from DNN, the essence of difference at frequency distribution. In this work, from the insight of frequency distribution, we rethink the SNN training process and give a novel method to transfer SNN to high-frequency spiking neural network (HF-SNN). This approach preserves considerably more information than other optimizing strategies and enables flexibility in the training process. Besides, we evaluate the HF-SNN with extensive experiments on three large datasets: CIFAR-10, CIFAR-100, and ImageNet. Finally, our model supports training a deeper SNN model from scratch and achieves better performance on these datasets than the existing SNN model.

We study communication systems over band-limited Additive White Gaussian Noise (AWGN) channels in which the transmitter’s output is constrained to be symmetric binary (bipolar). We improve the available Ozarow-Wyner-Ziv (OWZ) lower bound on capacity which is based on peak-power constrained pulse-amplitude modulation, by introducing new schemes (achievability) with two advantages over the studied OWZ schemes. Our schemes achieve a moderately improved information rate and they do so with much fewer sign transitions of the binary signal. The gap between the known upper bound, which is based on spectral constrains of bipolar signals, and our new achievable lower bound is reduced to 0.93 bits per Nyquist interval at high SNR.

Swarm Intelligence is applied to optimisation problems due to its robustness, scalability, generality, and flexibility. Based on simple rules, simple reactive agents - swarm (e.g. fish, bird, and ant) - directly or indirectly exchange information to find an optimal solution. Among multiple nature inspirations and versions, the dilemma of choosing proper swarm-based algorithms for each type of problem prevents their recurrent application. This scenario gets even more challenging when considering binary optimisation because of the absence of overview papers that assembles the trends, benefits and limitations of swarm-based techniques. Based on 403 scientific papers, we describe the basis of the leading binary swarm-based algorithms presenting their rationales, equations, pseudocodes, and descriptions of their applications to tackle this research gap. We also define a new classification based on the mechanism to update the solutions and the displacements, indicating that the Binary-Binary approach - binary decision variables and binary search space - is more efficient for binary optimisation in accuracy and computational cost.

Few-shot synthetic aperture radar (SAR) target classification has received more and more attention in recent years, where most of the existing methods have applied off-the-shelf networks designed for natural images to SAR images, ignoring the special characteristics of SAR data. Therefore, in this article, we propose an attribute-guided multi-scale prototypical network (AG-MsPN) combined with subband decomposition for few-shot SAR target classification, aiming to learn more discriminative features from a few labeled data. Since the SAR images are essentially complex-valued images containing both amplitude and phase information, we implement the subband decomposition of complex-valued SAR images to explore the backscattering variations of targets, thus obtaining more complete descriptions of targets. Then, considering the complementary features extracted by different convolutional layers, based on the prototypical network, a multi-scale prototypical network (MsPN) is proposed to fuse the features of different layers to enhance the discrimination of feature representations, thus relieving the problem of high intra-class diversity and inter-class similarity for the images of SAR targets. Besides, we devise the prior binary attributes of SAR targets and add an extra attribute classification module (ACM) into the MsPN to map the images into the attribute space for classification. During the training phase, the proposed MsPN and the ACM are jointly utilized to realize the target classification in both the feature space and the attribute space, and meanwhile, the model parameters are optimized by the joint loss. Thus, the classification performance of the MsPN is further enhanced under the joint supervision of class label information from a few labeled data and the target attribute information from the prior knowledge. Therefore, we name the proposed method the AG-MsPN. We demonstrate the effectiveness of our proposed AG-MsPN on the Moving and Stationary Target Acquisition and Recognition benchmark dataset, and our method surpasses many other existing methods in the few-shot cases.

This study proposes a novel binary optimization method for solving unit commitment (UC) problems. Developed from moth-flame optimization (MFO), the proposed method is called binary alternative MFO (BAMFO). Originally, each MFO moth is restricted to fly toward a predetermined flame (solution), causing it to be trapped in the local optimum of that flame and reducing its performance. To address these drawbacks, BAMFO represents four alternative characteristics for moths to discover better flames all over the search space. The exploration and exploitation performances of BAMFO via these characteristics are balanced by a random parameter competing with the criteria of each characteristic. A repair strategy working with the priority list is created to adjust any unit status from the BAMFO simulation to satisfy all UC constraints. Then, Lambda-iteration method is adopted to solve economic dispatch problems. The proficiency of BAMFO is verified in various scales of unit systems for short-term operational scheduling. The results show that BAMFO yields better solutions than do other methods. The Wilcoxon signed rank test was conducted to prove the predominance of BAMFO.

Recently, building an efficient and robust model for object detection has attracted the attention of the vision community. Although binary networks have a fast inference speed, they cannot be used directly on mobile devices such as unmanned aerial vehicles (UAVs) because of their low detection accuracy. Different from improving the detection accuracy of a binary network by adjusting the network structure or adjusting the update gradient, we propose an improved binary neural network based on the block scaling factor XNOR (BSF-XNOR) convolutional layer. In addition, we propose a two-level densely connected network structure, which further enhances the network layer’s feature representation capabilities. Experiments using the TensorFlow framework prove the effectiveness of our algorithm in improving accuracy. Compared with the original standard XNOR network, the mean average precision (mAP) detected by our algorithm on the PASCAL VOC dataset was improved. The experimental results on the VisDrone2019 UAV dataset confirm that our method achieves a better balance between speed and accuracy than previous methods. Our algorithm aims to guide and deploy high-precision binary networks on the embedded device and solves the problem of low-precision binary networks.

This paper proposes a depth from light field (DFLF) method specifically to deal with occlusion based on the foreground-background separation (FBS). The FBS-based methods infer the disparity maps by accumulating the binary maps which divide whether each pixel is the foreground or background. Although there have been widely studied to handle the occlusion problem with the cost-based method, there are not enough researches to handle the occlusion problem with the FBS-based methods yet. We found that errors around the occlusion boundary in the resulting disparity maps of the FBS-based methods arise from the fattened foreground by the light field reprameterization. To avoid fattened foregrounds, the inferred foreground maps in the front region with respect to the disparity axis could be utilized in the back region in the three-dimensional volume construction, which corresponds to the cost volume construction in the cost-based methods. With the front-to-back scanning manner of the FBS-based method, by successively excluding inferred foreground maps, errors around occlusion boundary could be effectively reduced in the resulting disparity maps. With synthetic and real LF images, the proposed method shows reasonable performance compared to the existing methods and better performance than existing FBS-based methods.

At present, there are many challenges in the field of pedestrian attribute recognition, such as small targets of some attributes, imbalanced samples, and low recognition accuracy of complex samples. In view of the above problems, we improved the model in two perspectives: 1) We proposed Feature Pyramid Attention Model (FPAM). In order to solve the problem that attributes are distributed in different locations in the pedestrian image, FPAM adopted the attention mechanism on the basis of ResNet-50, by which the model's attention could be focused on key areas of the image. As for the difficulty in small targets attributes, we adopted feature pyramid integration strategy; 2) We proposed Multi Label Focal Loss (MLFL). Referring to Binary Cross Entropy Loss Function (CE) and Weight Binary Cross Entropy Loss Function (WCE), we added the weight parameters of samples which are difficult to classify to improve the recognition accuracy, and the rate of convergence was increased. Results show that our proposed method achieves 84.83% mA, 79.37% Accuracy, 87.47% Precision, 86.09% Recall, and 86.77% F1 on PETA dataset.

In-memory computing (IMC) is a promising approach for energy cost reduction due to data movement between memory and processor for running data-intensive deep learning applications on the computing systems. Together with Binary Neural Network (BNN), IMC provides a viable solution for running deep neural networks at the edge devices with stringent memory and energy constraints. In this paper, we propose a novel 10T bit-cell with a back-end-of-line (BEOL) metal-oxide-metal (MOM) capacitor laid on pitch for in-memory computing. Our IMC bit-cell, when arranged in a memory array, performs binary convolution (XNOR followed by Bit-count operations) and binary activation generation operations. We show, when binary layers of BNN are mapped into our IMC arrays for MNIST digit classification, 98.75% accuracy with energy efficiency of 2193 TOPS/W and throughput of 22857 GOPS can be obtained. We determine the memory array size considering the word-line and bit-line nonidealities and show how these impact classification accuracy. We analyze the impact of process variations on classification accuracy and show how word-line pulse tunability provided by our design can be used to improve the robustness of classification under process variations.

The technique of binary code similarity detection (BCSD) has been applied in many fields, such as malware detection, plagiarism detection and vulnerability search, etc. Existing solutions for the BCSD problem usually compare specific features between binaries based on the control flow graphs of functions from binaries or compute the embedding vector of binary functions and solve the problem based on deep learning algorithms. In this paper, from another research perspective, we propose a new and lightweight method to solve cross-version BCSD problem based on multiple features. It transforms binary functions into vectors and signals and computes the similarity coefficient value and correlation coefficient value for solving cross-version BCSD problem. Without relying on the CFG of functions, deep learning algorithms and other related attributes, our method works directly on the raw bytes of each binary and it can be used as an alternative method to coping with various complex situations that exist in the real-world environment. We implement the method and evaluate it on a custom dataset with about 423,282 samples. The result shows that the method could perform well in cross-version BCSD field, and the recall of our method could reach 96.63%, which is almost the same as the state-of-the-art static solution.

Tri-state nanoelectromechanical (NEM) memory switches are proposed for the implementation of high-impedance state 0 in addition to low-impedance states 1 and 2 for the improvement of conventional complementary metal-oxide-semiconductor-NEM (CMOS-NEM) reconfigurable logic (RL) operations. Although it is well known that the high impedance state of routing switches is essential to prevent the unnecessary data throughput of RL circuits, previously proposed NEM memory switches have only implemented binary states: states 1 and 2. On the contrary, our proposed NEM memory switches can have tri-states, which are achieved by modifying their operation methods and design guidelines.

Recently, hybrid multiplication schemes over the binary extension field $GF(2^{m})$ based on $n$ -term Karatsuba algorithm (KA) have been proposed for irreducible trinomials. Their complexities depend on a decomposition of $m$ and the choice of a generation polynomial. However, these multipliers have some limitations on a decomposition of $m$ or generation polynomial $x^{m}+x^{k}+1$ such that $m\geq 2k$ . In this paper, we loosen such limited conditions. We present a new hybrid bit-parallel multiplier based on $n$ -term KA for any irreducible trinomial $x^{m}+x^{k}+1\,\,(0 , where $m$ is decomposed as $m=nm_{0}+r$ with $0 and $1 . (Here, various values for $n,m_{0}$ and $r$ may be chosen.) To this end, we generalize the previously proposed multiplication scheme for $x^{nm_{0}+1}+x^{k}+1$ into $x^{nm_{0}+r}+x^{k}+1$ . We evaluate the explicit complexity of the proposed multiplier. Specific comparisons show that the proposed multiplier achieves the lowest space complexity with the same or lower time complexity among hybrid multipliers. Compared to the fastest multipliers, the time complexity of the proposed multiplier costs only $T_{X}$ higher while its space complexity is much lower (it has roughly 40% reduced space complexity), where $T_{X}$ is the delay of one 2-input XOR gate.

Binary imbalance problem refers to such a classification scenario where one class contains a large number of samples while another class contains only a few samples. When traditional classifiers face with imbalanced datasets, they usually bias towards majority class resulting in poor classification performance. Oversampling is an effective method to address this problem, yet how to conduct diversity oversampling is a challenge. In this article, we proposed a diversity oversampling method based on a modified D2GAN model, and on the basis of diversity oversampling, we also proposed a binary imbalanced data classification approach based on classifier fusion by fuzzy integral. Extensive experiments are conducted on 8 data sets to compare the proposed methods with 7 state-of-the-art methods on 5 aspects: MMD-score, Silhouette-score, F-measure, G-means, and AUC-area. The 7 methods include 4 SMOTE related approaches and 3 GAN related approaches. The experimental results demonstrate that the proposed methods are more effective and efficient than the compared approaches.

A time sharing (TS) hybrid probabilistic shaping (HPS) scheme for nonbinary (NB) LDPC codes is proposed. For the conventional probabilistic shaping (CPS) scheme, the demodulator uses bit-metric decoding (BMD) for binary LDPC codes and NB LDPC codes. However, the demodulator for NB LDPC codes adopts symbol-metric decoding (SMD) on the AWGN channel to avoid the loss of symbol-to-bit mapping and obtain better performance than binary LDPC codes. In order to take advantage of SMD demodulation for NB LDPC codes, the NB-LDPC information symbols use constant composition distribution matcher (CCDM) to generate quadrature amplitude modulation (QAM) symbols with unequal probability distribution, while the NB-LDPC parity symbols are utilized to yield QAM symbols with approximately uniform distribution. The SMD demodulation of the proposed NB-LDPC-coded TS HPS scheme has lower demodulation complexity than the NB-LDPC-coded CPS scheme. The information-theoretical analysis of the CPS system and the TS HPS system turns out that the proposed NB-LDPC-coded TS HPS scheme is superior to the NB-LDPC-coded CPS scheme for 8QAM and 32QAM format. Simulation results verify the theoretical average mutual information (AMI) analysis and show that the proposed NB-LDPC-coded TS HPS scheme has up-to-0.24 dB and 1.10 dB performance gain compared with the NB-LDPC-coded CPS scheme for 8QAM and 32QAM format, respectively. Compared with the binary LDPC-coded regular Nyquist scheme for 8QAM and 32QAM format, the proposed NB-LDPC-coded TS HPS scheme has up-to-1.31 dB and 2.04 dB performance gain, respectively. Meanwhile, the LDPC code rate of the proposed TS HPS scheme is more flexible than the CPS scheme. Therefore, the proposed NB-LDPC-coded TS HPS scheme is more reliable, efficient, and flexible for 8QAM and 32QAM, and thus suitable for the sixth generation (6G) communication system.

We describe properties and constructions of constraint-based codes for DNA-based data storage which account for the maximum repetition length and AT/GC balance. Generating functions and approximations are presented for computing the number of sequences with maximum repetition length and AT/GC balance constraint. We describe routines for translating binary runlength limited and/or balanced strings into DNA strands, and compute the efficiency of such routines. Expressions for the redundancy of codes that account for both the maximum repetition length and AT/GC balance are derived.

The search for a binary switch that is more energy-efficient than a transistor has led to many ideas, notable among which is the notion of using a nanomagnet with two stable magnetization orientations that will encode the binary bits 0 and 1. The nanomagnet is switched between them with electrically generated mechanical strain. A tiny amount of voltage is required for switching, with energy dissipation on the order of a few to few tens of aJ. Logic gates and memory, predicated on this technology, have been demonstrated in our group. While they indeed dissipate very little energy, they are unfortunately plagued by unacceptably high switching error probability that hinders their application in most types of Boolean logic. Fortunately, they can still be used in applications that are more forgiving of switching errors, e.g. probabilistic computing, analog arithmetic circuits, belief networks, artificial neurons, restricted Boltzmann machines, image processing, and others where the collective activity of many devices acting cooperatively elicit the computing or signal processing function and the failure of a single or few devices does not matter critically. These ultra-energy-efficient strain-switched nanomagnets can also be used for non-computing devices such as microwave oscillators that perform better than spin-torque-nano-oscillators. This short review provides an introduction to this exciting burgeoning field.

Orthogonal frequency division multiplexing (OFDM) has been widely used in multi-carrier communication systems. However, high the peak-to-average power ratio (PAPR) is a very troublesome problem in OFDM system. In the paper, we propose a construction of Golay complementary sets (GCSs) with size 4 of flexible lengths using even-shift complementary pairs (ESCPs). Based on computer search, we find that there exist a large number of even-shift complementary pairs of various lengths, and we also propose some constructions of ESCPs, which are very useful for the construction of GCSs.

In this work, the low-density parity-check (LDPC)-coded modulation scheme which has been selected for inclusion in the next generation of International Telecommunication Union (ITU) broadband standard is investigated. The multi-level mapping of this scheme offers excellent performance along with a straightforward modulation which is separable into simple one-dimensional schemes. A subset of modulated bit positions are protected by the LDPC code, while the remaining positions benefit from improved Euclidean distance through assignment to distinct constellation regions. To date, there has been a lack of analytical treatment of the scheme. This is problematic given the number of tunable scheme and code parameters, necessitating time-consuming Monte Carlo simulation. This motivates the novel analytical work of this article, which derives overall error-rate and threshold performances through considering separately the effects of the channel on the coded and uncoded bits. First, new closed-form expressions are derived for the hard-decision performance of the bit mapping of the multi-level modulation scheme, for both coded and uncoded bits. Extrinsic information transfer (EXIT) analysis is then applied to the coded bits in the scheme. The derived theoretical performance of the uncoded bit positions is used in combination with the derived EXIT threshold to provide for the first time for this modulation scheme a method to evaluate analytically whether a designed system will offer capacity-approaching performance. Following this, an approximation to the capacity curve for the coded modulation scheme is produced, again based on the derived analytical performance. Finally, through the relationship between mutual information and error rate performance for error control codes, the derived analytical expressions are used to produce a semi-analytical finite-length performance predictor, whereby error rate results for a given code on the binary input additive white Gaussian noise (AWGN) channel can be mapped to the coded modulation scheme under consideration for any coded system parameters.

Preservation of image features caused by binary conversion is a difficult task under variation of illumination conditions. Several binary conversion-based methods have used an adaptive thresholding technique to improve their performance under illumination variation conditions because of its robustness. However, the performances of existing methods were still limited under high differences illumination conditions especially for uncontrolled lighting sources. In addition, various length of face-to-camera distance gives significant problem affect for the performance of face recognition method. It happens when various images are available for the same person with different length face-to-camera distances due to the appearance of varying facial features of the same person. Therefore, this study proposed to combine the strength normalization and feature-based method to build an illumination distribution model to overcome this problem. With the proposed method the illumination model will fit with variation of illumination conditions in a whole image to generate an adaptive threshold for a novel columnar binary conversion method. The proposed method consists of five main stages, starting with eye area detection using the developed Viola–Jones algorithm. Next, the iris is detected using the Circular Hough Transform (CHT) method and will convert it into binary using the proposed Columnar Binary Conversion (CBC) method to preserve the appearance of the facial features under the illumination variation. Then, the proposed Facial Feature Region Normalization (FFRN) method is performed to improve the effects of different optical zooms for the classification step. The classification is conducted based on the similarity measurement between the extracted normalised binary face region and the dataset that must be converted into their equivalent normalised binary images. The proposed method is evaluated on two different smartphone databases, namely as Smartphone Face Video (SFV) and MOBIO. The performance results showed the outperformance of the proposed method.

Diatoms are currently being used in different fields, and their application as bioindicators is one of their main benefits. Their identification is carried out by specialists, the main difficulty being the wide variety of existing species and the great similarity between some of them. In recent years attempts have been made to create an automatic identification system with the collaboration of experts in areas of pattern recognition. In this research we analyzed the use of binary masks of rings in images as a method to automatically identify diatoms, or render an approximative identification. Among the advantages of the proposed method is its invariance regarding the different positions in which the diatoms can be found. Invariance to illumination changes is achieved by means of different transforms, such as the power law transform. The image databases used to develop the algorithms were diatom images taken from samples collected at springs in Lake Patzcuaro, Michoacan, Mexico, at the San Lazaro basin, Baja California Sur, Mexico and from the public image database of ADIAC. In morphological and textural descriptors, the background of the image must be eliminated in order to classify the diatoms, whereas in the proposed methodology this is not necessary. The obtained results reach over than 90% even when the diatoms in the images are broken.

Multi-label classification tackles the problems in which each instance is associated with multiple labels. Due to the interdependence among labels, exploiting label correlations is the main means to enhance the performances of classifiers and a variety of corresponding multi-label algorithms have been proposed. Among those algorithms Classifier Chains (CC) is one of the most effective methods. It induces binary classifiers for each label, and these classifiers are linked in a chain. In the chain, the labels predicted by previous classifiers are used as additional features for the current classifier. The original CC has two shortcomings which potentially decrease classification performances: random label ordering, noise in original and additional features. To deal with these problems, we propose a novel and effective algorithm named LSF-CC, i.e. Label Specific Features based Classifier Chain for multi-label classification. At first, a feature estimating technique is employed to produce a list of most relevant features and labels for each label. According to these lists, we define a chain to guarantee that the most frequent labels that appear in these lists are top-ranked. Then, label specific features can be selected from the original feature space and label space. Based on these label specific features, corresponding binary classifiers are learned for each label. Experiments on several multi-label data sets from various domains have shown that the proposed method outperforms well-established approaches.

This article addresses the question of implementing a maximum flow algorithm on directed graphs in a formulation suitable for a quantum annealing computer. Three distinct approaches are presented. In all three cases, the flow problem is formulated as a quadratic unconstrained binary optimization (QUBO) problem amenable to quantum annealing. The first implementation augments a graph with integral edge capacities into a multigraph with unit-capacity edges and encodes the fundamental objective and constraints of the maximum flow problem using a number of qubits equal to the total capacity of the graph $\sum _i{c_i}$ . The second implementation, which encodes flows through edges using a binary representation, reduces the required number of qubits to $\mathcal {O}(|E| \log C_{\max })$ , where $|E|$ and $C_{\max }$ denote the number of edges and maximum edge capacity of the graph, respectively. The third implementation adapts the dual minimum cut formulation and encodes the problem instance using $|V|$ qubits, where $|V|$ is the number of vertices in the graph. Scaling factors for penalty terms and coupling matrix construction times are made explicit in this article.

In this work, we study two classes of quasi-cyclic (QC) codes and examine how several properties can be combined into the codes of these classes. We start with the class of QC codes generated by diagonal generator polynomial matrices; a QC code in this class is a direct sum of cyclic codes. Then we move on to the class of QC codes of index 2; various binary codes with good parameters are found in this class. In each class, we describe the generator polynomial matrices of reversible codes, self-orthogonal codes, and self-dual codes. Hence, we demonstrate how such properties can be merged in codes of these classes. Particularly for QC codes of index 2, we prove a necessary and sufficient condition for the self-orthogonality of reversible codes. Then we show that reversible QC codes of index 2 are self-dual under the same conditions in which self-dual codes are reversible. We clarify that self-orthogonal reversible QC codes of index 2 over $\mathbb {F}_{q}$ exist for even and odd $q$ , however self-dual reversible codes exist only for even $q$ . Theoretical results are reinforced by several numerical examples. Computer search is used to present some self-dual reversible QC codes of index 2 that have the best known parameters as linear codes. Finally, we highlight the class of 1-generator binary QC codes of index 2 by exploring many self-dual reversible codes that achieve the upper bound on the minimum distance for their parameters.

Recently, new multiplication schemes over the binary extension field $GF(2^{m})$ based on an $n$ -term Karatsuba algorithm have been proposed for irreducible trinomials. In this paper, we extend these schemes for trinomials to any irreducible polynomials. We introduce some new types of pentanomials and propose multipliers for those pentanomials utilizing the extended schemes. We evaluate the rigorous space and time complexities of the proposed multipliers, and compare those with similar bit-parallel multipliers for pentanomials. As a main contribution, the best space complexities of our multipliers are $\frac {1}2m^{2}+O\left({m^{\frac {3}2}}\right)$ AND gates and $\frac {1}2m^{2}+O\left({m^{\frac {3}2}}\right)$ XOR gates, which nearly correspond to the best results for trinomials. Also, specific comparisons for three fields $GF(2^{163})$ , $GF(2^{283})$ , and $GF(2^{571})$ recommended by NIST show that the proposed multiplier has roughly 40% reduced space complexity compared to the fastest multipliers, while it costs a few more XOR gate delay. It is noticed that our space complexity gain is much greater than the time complexity loss. Moreover, the proposed multiplier has about 21% reduced space complexity than the best-known space efficient multipliers, while having the same time complexity. The results show that the proposed multipliers are the best space optimized multipliers.

Although a number of change detection (CD) methods have been proposed during the past years, most of them are developed based on the assumption that there are either training samples or no training samples for both the pretime and posttime images. Few studies have addressed the scenario of only small amounts of training samples are available only in a single-time image. In this article, we propose a novel approach that can detect multiple changes in bitemporal hyperspectral images when only a few training samples are available in one of the images (the source image). The proposed method consists of four main steps: first, unsupervised CD based on uncertain area analysis to generate the binary change map; second, classification of the source image (X1) according to active learning; third, classification of the target image (X2) by the use of improved transfer learning; and fourth, generation of the multiclass change map by postclassification comparison. The proposed method was tested on one simulated dataset and two pairs of real bitemporal hyperspectral images. Experimental results demonstrate that: first, uncertain area analysis can improve the binary CD accuracy; while active learning and improved transfer learning can enhance the classification accuracy of the source and target images, the multiple CD accuracy is increased by the use of the proposed method; and second, compared with the state-of-the-art methods, the proposed method produced best results.

This paper proposes a multifocal lens design method named phase superposition method (PSM) and a methodology to increase the efficiency of metallic binary zone plates through a sub-wavelength slit array. The phase superposition method calculates the intensity at each focus, assuming a perfectly phase-matched modulation based on input light, and the size of the lens. Subsequently, based on the calculated focus intensity, the phases linearly related to the target intensity at each focus are combined. The sub-wavelength slit-assisted metallic binary plate assumes the sub-wavelength slit array as a dielectric in the light polarized perpendicular to the slit. It then controls the condition of the slit to invert the phase difference between the opened region and the slit region. By applying these methods, the trifocal lens was designed to verify our idea. The light was focused on three target positions. After applying the sub-wavelength slit structure, the amplitude was increased up to 90.11%, and the efficiency increase was 303.99% compared to that of a binary plate without the sub-wavelength slit.

Feature selection is an effective method to eliminate irrelevant, redundant and noisy features, which improves the performance of classification and reduces the computational burden in machine learning. In this paper, an improved binary dragonfly algorithm (IBDA) which extends from the conventional dragonfly algorithm (DA) is proposed as a search strategy to design a wrapper-based feature selection method. First, a novel evolutionary population dynamics (EPD) strategy is introduced in IBDA to enhance the exploitation ability while ensuring population diversity of the algorithm. Second, IBDA proposes a novel crossover operator which establishes connections between the crossover rates and iterations so that making the algorithm can adjust the crossover rates of solutions dynamically, thereby balancing the exploitation and exploration of the algorithm. Finally, a binary mechanism is proposed to make the algorithm suitable for the binary feature selection problems. Simulations are conducted on 27 classical datasets from the UC Irvine Machine Learning Repository, and the results demonstrate that the proposed IBDA has better performance than some other comparison algorithms. Moreover, the effectiveness and performance of the proposed improved factors are evaluated by tests.

A photonic method to generate hybrid binary digitally modulated radio frequency (RF) signals (i.e. ASK, PSK, and FSK) by using a polarization-division-multiplexed Mach-Zehnder modulator (PDM-MZM) is proposed. The PDM-MZM is driven by two RF signals with different carrier frequencies and two multiplied digital signals. By properly controlling the offsets and the amplitudes of the digital signals, ASK, PSK, and FSK signals can be generated and switched fastly. RF signals at 4/6 GHz and 10/15 GHz with switchable modulation formats are experimentally generated. Moreover, the performances of the generated signals are evaluated in a wireless transmission system and corresponding bit error rate (BER) curves of the signals are presented.

NTRUEncrypt is a public key cryptosystem based on hard problems over lattices. The dominant operation in NTRUEncrypt is convolution, i.e., multiplication over a quotient ring of polynomials. Based on the fact that a convolution has a highly regular structure, Lee et al. proposed the sliding window method for fast convolution of binary polynomials in 2013, which was then extended to ternary polynomials for ideal lattices by Akleylek, Alkim, and Tok in 2016. These sliding window methods reduce the cost of a convolution operation using look-up tables that store partial computation results related to repeated coefficient patterns. In this paper, we propose a signed sliding window method with side-channel resistance for NTRUEncrypt. The proposed method considers both positive and negative nonzero coefficients when constructing look-up tables. The new method not only accelerates convolution but also enables the application of power analysis countermeasures effectively. According to the experimental results, the constant-time implementation of the proposed method with timing and power analysis countermeasures accelerates the previously developed secure convolution method by up to 20%.

The majority of existing few-shot learning methods describe image relations with binary labels. However, such binary relations are insufficient to teach the network complicated real-world relations, due to the lack of decision smoothness. Furthermore, current few-shot learning models capture only the similarity via relation labels, but they are not exposed to class concepts associated with objects, which is likely detrimental to the classification performance due to underutilization of the available class labels. To paraphrase, children learn the concept of tiger from a few of actual examples as well as from comparisons of tiger to other animals. Thus, we hypothesize that in fact both similarity and class concept learning must be occurring simultaneously. With these observations at hand, we study the fundamental problem of simplistic class modeling in current few-shot learning methods. We rethink the relations between class concepts, and propose a novel Absolute-relative Learning paradigm to fully take advantage of label information to refine the image representations and correct the relation understanding in both supervised and unsupervised scenarios. Our proposed paradigm improves the performance of several the state-of-the-art models on publicly available datasets., IEEE/CVF Conference on Computer Vision and Pattern Recognition 2021

The recent advance in the post-quantum cryptography (PQC) field has gradually shifted from the theory to the implementation of the cryptosystem, especially on the hardware platforms. Following this trend, in this paper, we aim to present efficient implementations of the finite field arithmetic (key component) for the binary Ring-Learning-with-Errors (Ring-LWE) PQC through a novel lookup-table (LUT)-like method. In total, we have carried out four stages of interdependent efforts: (i) an algorithm-hardware co-design driven derivation of the proposed LUT-like method is provided detailedly for the key arithmetic of the BRLWE scheme; (ii) the proposed hardware architecture is then presented along with the internal structural description; (iii) we have also presented a novel hybrid size structure suitable for flexible operation, which is the first report in the literature; (iv) the final implementation and comparison processes have also been given, demonstrating that our proposed structures deliver significant improved performance over the state-of-the-art solutions. The proposed designs are highly efficient and are expected to be employed in many emerging applications.

Approximate computing can increase performance, reduce power and energy consumption for applications that can tolerate computation inaccuracy. Current commodity computers use predefined bit-width, typically from 8- to 64-bit binary to represent the data. Unfortunately, the data precision required is application specific and current computers do not support specific data precision for a specific application. We propose a bit-serial architecture to support reconfigurable bit-width. The bit-width can be configured for varied computation accuracy, from full accuracy to any number of bits to represent the data to trade off computation accuracy and output quality for performance, power, and energy. The proposed bit-serial architecture is emulated and evaluated on a conventional commodity computer that only supports fixed bitwidth computations. We use an image processing application to demonstrate the benefits of bit-serial architecture. The speedup is from 11.6x to 3.6x and energy consumption reduction is from 8.2x to 2.5x when bit-width is from 3- to 7-bit compared with a bit-parallel baseline.

Binary and a ternary message passing decoding algorithms for generalized low-density parity-check codes are defined and analyzed. In all algorithms, the variable nodes exploit soft channel information. However, the exchanged messages between variable and check nodes are binary or ternary. Two types of local decoders are assumed at the check nodes, namely an optimum a-posteriori probability soft-input soft-output decoder, and a more practical bounded distance decoder. A density evolution analysis for irregular code ensembles is developed for all the introduced algorithms, which allows obtaining the scaling coefficients needed for the variable node operation. Finite-length simulation results confirm the validity of the asymptotic analysis.

The study of robustness of deep classifiers has exposed their vulnerability to perturbation attacks. Prior work has largely focused on adversarial attacks targeting one-stage-classifiers. By contrast, here we investigate the susceptibility of Nested Dichotomies Classifiers (NDCs), which decompose a multiclass problem into a collection of binary ones, to such types of individual attacks. First, we show that the overall regret of an NDC is the sum of regrets of the binary classifiers along the path from the root to the leaf nodes of these dichotomies. Then, we formulate an optimization program to generate perturbations fooling NDCs and propose an algorithmic solution based on a convex relaxation. A solution is obtained by developing an ADMM-based solver to the convex programs. The experiments show that NDCs are more robust than their single stage counterpart in that the optimal perturbations inducing misclassifications are more perceptible.