Embedding Phase Research Articles

Low-complexity coherent DSCM system with precise transceiver IQ skew calibration and simplified equalization embedding phase tracking.

The digital subcarrier multiplexing (DSCM) technique has emerged as an appealing solution for coherent systems. However, it is more sensitive to the transceiver IQ skew, which introduces conjugate interference to the symmetric frequency subcarriers (SC). In this Letter, a low-complexity transceiver IQ skew calibration method is adopted for coherent DSCM systems based on a specially designed training signal, avoiding the power-hungry multiple-input multiple-output (MIMO) operation between two symmetric SCs. To further reduce the computational complexity for each SC, a simplified equalizer with embedded phase tracking (SEQPT) is proposed, achieving polarization de-multiplexing, channel equalization, and phase noise compensation simultaneously. The proposed methods are validated experimentally in a 16-channel wavelength division multiplexing system with 50 GBd 4SC-16QAM DSCM signals. The results indicate that after skew calibration by implementing the proposed method, performance is comparable to that of applying the 8 × 8 real-valued MIMO equalizer. Besides, using the proposed SEQPT incurs almost no performance penalty compared to the N-tap 2 × 2 CMMA equalizer followed by blind phase search while operating in a low-complexity manner.

An Efficient Cryptographic Scheme based on Optimized Watermarking Scheme for Securing Internet of Things

In this work, a new efficient cryptographic scheme based on the concept of chaotic map and optimized watermarking scheme is proposed. In the optimized watermarking scheme, a combination of discrete wave transformation (DWT), hessenberg decomposition (HD), and singular value decomposition (SVD) are used. In this, the host image is first broken down into several sub-bands using multi-level DWT, and the resulting coefficients are then fed into HD during the embedding phase. Simultaneous watermark operation is performed on SVD. Finally, the scale factor embeds the watermark into the host image. The Differential evolution method is used to find the best scaling factor for the optimized watermarking scheme. The resulting watermarked image is then encrypted by the session key based scheme. In this scheme for each image encryption, a new random session key will be produced. The presented approach uses 64-bit plaintext and a variable size key that will be decided at the time of encryption for encrypting an image. Since session keys change with each transmission, this approach does not involve extracting and remembering session keys in order to produce subsequent session keys. IoT devices are used to test the developed method for security. The experiment’s findings shows that the suggested method works better than the current scheme in several aspects.

Improvement of the Thermal Performance of PCM-Based Heat Sink Used in Electronic Cooling by Adding Nano-Particles

Recently, thanks to the technological advances, electronic devices are getting smaller in size. This causes an increase in the heat generation per unit area. This heat has to be removed from electronic devices for them to be longer-lasting, more efficient and more reliable. There are many active and passive methods designed for this objective. One of them is embedding phase change material (PCM) in the heat sink. PCM, during the phase change stage, absorbs the heat generated in the system and thus aids in keeping the temperature at a certain value. The biggest downside of PCM is its rapid conduction of heat. PCM properties can be improved by using nanoparticles. In this study, nanoparticles such as TiO2 and CuO were added to PCM and such a modified PCM is used in a finned heat sink. The thermal behavior of the PCM with addition of 1%, 2% and 5% TiO2 and CuO was investigated numerically in three dimensions. RT-28HC was used as the PCM in the study. It was shown that as the nanoparticle ratio increases, heat transfer coefficient of the PCM rises and the melting time of Nanoparticle PCM (NPPCM) is less than that of pure PCM. However, it was observed that, the melting time of PCM with CuO added is longer than that of the PCM with TiO2 added.

Unambiguous and High-Fidelity Backdoor Watermarking for Deep Neural Networks.

The unprecedented success of deep learning could not be achieved without the synergy of big data, computing power, and human knowledge, among which none is free. This calls for the copyright protection of deep neural networks (DNNs), which has been tackled via DNN watermarking. Due to the special structure of DNNs, backdoor watermarks have been one of the popular solutions. In this article, we first present a big picture of DNN watermarking scenarios with rigorous definitions unifying the black- and white-box concepts across watermark embedding, attack, and verification phases. Then, from the perspective of data diversity, especially adversarial and open set examples overlooked in the existing works, we rigorously reveal the vulnerability of backdoor watermarks against black-box ambiguity attacks. To solve this problem, we propose an unambiguous backdoor watermarking scheme via the design of deterministically dependent trigger samples and labels, showing that the cost of ambiguity attacks will increase from the existing linear complexity to exponential complexity. Furthermore, noting that the existing definition of backdoor fidelity is solely concerned with classification accuracy, we propose to more rigorously evaluate fidelity via examining training data feature distributions and decision boundaries before and after backdoor embedding. Incorporating the proposed prototype guided regularizer (PGR) and fine-tune all layers (FTAL) strategy, we show that backdoor fidelity can be substantially improved. Experimental results using two versions of the basic ResNet18, advanced wide residual network (WRN28_10) and EfficientNet-B0, on MNIST, CIFAR-10, CIFAR-100, and FOOD-101 classification tasks, respectively, illustrate the advantages of the proposed method.

Adaptive 3D Mesh Steganography Based on Feature-Preserving Distortion.

Current 3D mesh steganography algorithms relying on geometric modification are prone to detection by steganalyzers. In traditional steganography, adaptive steganography has proven to be an efficient means of enhancing steganography security. Taking inspiration from this, we propose a highly adaptive embedding algorithm, guided by the principle of minimizing a carefully crafted distortion through efficient steganography codes. Specifically, we tailor a payload-limited embedding optimization problem for 3D settings and devise a feature-preserving distortion (FPD) to measure the impact of message embedding. The distortion takes on an additive form and is defined as a weighted difference of the effective steganalytic subfeatures utilized by the current 3D steganalyzers. With practicality in mind, we refine the distortion to enhance robustness and computational efficiency. By minimizing the FPD, our algorithm can preserve mesh features to a considerable extent, including steganalytic and geometric features, while achieving a high embedding capacity. During the practical embedding phase, we employ the Q-layered syndrome trellis code (STC). However, calculating the bit modification probability (BMP) for each layer of the Q-layered STC, given the variation of Q, can be cumbersome. To address this issue, we design a universal and automatic approach for the BMP calculation. The experimental results demonstrate that our algorithm achieves state-of-the-art performance in countering 3D steganalysis.

First order phase transition in the D3-D7 model from the point of view of the fermionic spectral functions

We consider the D3-D7 model and use the spectral function of a probe fermion on D7 to analyze the first order phase transition from the black-hole embedding phase to another black-hole embedding phase in the presence of the finite density and temperature. From the fermionic spectral functions, we study the temperature dependence of the decay rate, and we observe various phenomena that support the first order phase transition including jump in it at the critical temperature that corresponds to the first order phase transition. Published by the American Physical Society 2024

Physics-constrained robust learning of open-form partial differential equations from limited and noisy data

Unveiling the underlying governing equations of nonlinear dynamic systems remains a significant challenge. Insufficient prior knowledge hinders the determination of an accurate candidate library, while noisy observations lead to imprecise evaluations, which in turn result in redundant function terms or erroneous equations. This study proposes a framework to robustly uncover open-form partial differential equations (PDEs) from limited and noisy data. The framework operates through two alternating update processes: discovering and embedding. The discovering phase employs symbolic representation and a novel reinforcement learning (RL)-guided hybrid PDE generator to efficiently produce diverse open-form PDEs with tree structures. A neural network-based predictive model fits the system response and serves as the reward evaluator for the generated PDEs. PDEs with higher rewards are utilized to iteratively optimize the generator via the RL strategy and the best-performing PDE is selected by a parameter-free stability metric. The embedding phase integrates the initially identified PDE from the discovering process as a physical constraint into the predictive model for robust training. The traversal of PDE trees automates the construction of the computational graph and the embedding process without human intervention. Numerical experiments demonstrate our framework's capability to uncover governing equations from nonlinear dynamic systems with limited and highly noisy data and outperform other physics-informed neural network-based discovery methods. This work opens new potential for exploring real-world systems with limited understanding.

Influential Metrics Estimation and Dynamic Frequency Selection Based on Two-Dimensional Mapping for JPEG-Reversible Data Hiding.

JPEG Reversible Data Hiding (RDH) is a method designed to extract hidden data from a marked image and perfectly restore the image to its original JPEG form. However, while existing RDH methods adaptively manage the visual distortion caused by embedded data, they often neglect the concurrent increase in file size. In rectifying this oversight, we have designed a new JPEG RDH scheme that addresses all influential metrics during the embedding phase and a dynamic frequency selection strategy with recoverable frequency order after data embedding. The process initiates with a pre-processing phase of blocks and the subsequent selection of frequencies. Utilizing a two-dimensional (2D) mapping strategy, we then compute the visual distortion and file size increment (FSI) for each image block by examining non-zero alternating current (AC) coefficient pairs (NZACPs) and their corresponding run lengths. Finally, we select appropriate block groups based on the influential metrics of each block group and proceed with data embedding by 2D histogram shifting (HS). Extensive experimentation demonstrates how our method's efficiently and consistently outperformed existing techniques with a superior peak signal-to-noise Ratio (PSNR) and optimized FSI.

End-to-End Phase Field Model Discovery Combining Experimentation, Crowdsourcing, Simulation and Learning

The availability of tera-byte scale experiment data calls for AI driven approaches which automatically discover scientific models from data. Nonetheless, significant challenges present in AI-driven scientific discovery: (i) The annotation of large scale datasets requires fundamental re-thinking in developing scalable crowdsourcing tools. (ii) The learning of scientific models from data calls for innovations beyond black-box neural nets. (iii) Novel visualization & diagnosis tools are needed for the collaboration of experimental and theoretical physicists, and computer scientists. We present Phase-Field-Lab platform for end-to-end phase field model discovery, which automatically discovers phase field physics models from experiment data, integrating experimentation, crowdsourcing, simulation and learning. Phase-Field-Lab combines (i) a streamlined annotation tool which reduces the annotation time (by ~50-75%), while increasing annotation accuracy compared to baseline; (ii) an end-to-end neural model which automatically learns phase field models from data by embedding phase field simulation and existing domain knowledge into learning; and (iii) novel interfaces and visualizations to integrate our platform into the scientific discovery cycle of domain scientists. Our platform is deployed in the analysis of nano-structure evolution in materials under extreme conditions (high temperature and irradiation). Our approach reveals new properties of nano-void defects, which otherwise cannot be detected via manual analysis.

Virtual network embedding method based on node delay perception

AbstractWireless virtual network uses software defined network and network function virtualisation technologies to create multiple logically isolated virtual networks on a physical wireless network. Wireless virtual network can improve the utilisation of wireless resources to meet the requirements of different services. Delay is an important performance indicator, which has strict requirements for delay‐sensitive services such as video conferencing and online games. In this study, a virtual network embedding method based on node delay perception (VNE‐NDP) is proposed, which considers both the node and link resources as well as the embedding delay requirements of virtual networks. The virtual network embedding method based on node delay perception consists of two phases: virtual node embedding and virtual link embedding. In the virtual node embedding phase, a physical node sorting method based on node delay perception (PNS‐NDP) is proposed. In PNS‐NDP, the node deployment delay is introduced into the node sorting algorithm for the first time. The authors select candidate physical nodes for each virtual node according to their resource availability and delay performance, which can greatly reduce the VN embedding delay without sacrificing too much other performance. In the virtual link embedding phase, a shortest path algorithm with bandwidth and link deployment delay constraints to find feasible physical paths for each virtual link is used. In addition, the VN embedding (VNE) deployment time is set as a new evaluation index. Simulation results show that compared with other VNE methods, VNE‐NDP can achieve higher success rate, revenue‐to‐expenditure ratio, and lower deployment delay.

Blind image watermark decoder in NSST-FPCET domain using Weibull Mixtures-HMT

Balancing imperceptibility, robustness, and data payload is a key topic in digital watermarking technology. Many contributions point to statistical modeling as an effective solution to this problem. On the basis of this, in this paper, we devise a hybrid domain image watermark decoder based on Weibull mixtures based hidden Markov tree (Weibull Mixtures-HMT) model. In the embedding phase, Otsu-Canny edge detection method is used to map the positions of high entropy blocks to the target subband obtained by non-subsampled shearlet transform (NSST), and fast polar complex exponential transform (FPCET) is computed in the target blocks to obtain the NSST-FPCET domain. The watermark signals are embedded into NSST-FPCET magnitudes by a linear method. In the extraction phase, NSST-FPCET magnitude coefficients are modeled by Weibull Mixtures-HMT model, which describes the distribution characteristics as well as the dependencies of NSST-FPCET magnitudes. The efficient variance reduced stochastic expectation maximization method is employed for estimating the parameters of Weibull Mixtures-HMT model. A statistical image watermark decoder in NSST-FPCET domain is finally achieved by the maximum likelihood decision. Massive experiments are performed so as to confirm the superiority of the designed scheme in trade-off among the data payload, imperceptibility, and robustness.

Neural-based inexact graph de-anonymization

Graph de-anonymization is a technique used to reveal connections between entities in anonymized graphs, which is crucial in detecting malicious activities, network analysis, social network analysis, and more. Despite its paramount importance, conventional methods often grapple with inefficiencies and challenges tied to obtaining accurate query graph data. This paper introduces a neural-based inexact graph de-anonymization, which comprises an embedding phase, a comparison phase, and a matching procedure. The embedding phase uses a graph convolutional network to generate embedding vectors for both the query and anonymized graphs. The comparison phase uses a neural tensor network to ascertain node resemblances. The matching procedure employs a refined greedy algorithm to discern optimal node pairings. Additionally, we comprehensively evaluate its performance via well-conducted experiments on various real datasets. The results demonstrate the effectiveness of our proposed approach in enhancing the efficiency and performance of graph de-anonymization through the use of graph embedding vectors.

What are the mechanisms driving the early stages of embedded researcher interventions? A qualitative process evaluation in English local government

Improved collaboration and communication between public health practitioners and academia could enhance the flow of research evidence into policy and practice. Embedded researchers present one type of intervention with the potential to bridge the research-implementation gap through their dual affiliations with decision makers and academia. Although embedded researcher posts are garnering increasing attention in public health, there remains a need to understand the mechanisms through which they may promote the translation of evidence into practice. To address this gap, we conducted a processes evaluation incorporating data from seventeen semi-structured interviews with embedded researchers in local government public health teams across England. We aimed to expand theoretical understandings of embedded researchers in public health through providing a detailed conceptualisation of the mechanisms shaping the early stages of their roles. Interviews with embedded researchers were conducted from late 2021 to spring 2022. Our results suggest that the initial months of embedded researcher roles are defined by a lengthy embedding phase centred on building trust and gathering contextual knowledge. This phase forms the foundation on which these interventions are built. We identified seven categories of outputs delivered by embedded researchers which primarily revolved around building research capacity and addressed many of the primary barriers limiting research activity in public health. Improvements in research awareness, interest, and involvement reflected early changes in local research cultures. However, our results align with previous work suggesting that changing an organisational research culture is a long-term process. Expectations for embedded researchers should thus be proportionate to the seniority and scale of the post and we add our voice to calls for sustained investment in these valuable interventions. Further examination of how embedded researcher roles evolve over time in public health is necessary to broaden understandings of the concept of embeddedness in these settings.

Embedding phase reduction for fast-slow systems with noise-induced stochastic quasiperiodic orbits

Noise is ubiquitous in natural systems and can induce coherent stochastic oscillations which cannot occur in the absence of noise, or stochastic quasiperiodic orbits which are markedly different from the deterministic limit cycle. For fast-slow systems, noise acting on the fast subsystem can induce quasi-periodic orbits tuned by the slow variable, which were termed as self-induced stochastic resonance (SISR). In our previous work [Zhu et al., 2022 [46]], we proposed a phase reduction framework on the SISR-oscillator by the hybrid approximation on the stochastic quasiperiodic orbit and the estimation of the effective noise. In this work, we remove the restriction of the additive Gaussian form assumption on the effective noise, and propose a more natural dimensionality-reduction method which we refer to as embedding phase reduction. Different SISR-oscillators induced by varying noise intensities, which can be embedded into a limit cycle, can have a unified form of the reduced phase equation. The efficiency of the embedding phase reduction is validated by Monte Carlo simulations of the case of periodic forcing, where excellent agreement is achieved for the entrainment and phase drifting dynamics. The simplicity and generality of the proposed embedding phase reduction approach indicate its potential applications in practical experiments and relevance in biological systems.

Deep learning-based image watermarking technique with hybrid DWT-SVD

ABSTRACT In digital media copyright protection, image watermarking topologies afford a promising solution. But the robustness of watermarking methods should be considered. Therefore, a robust image watermarking technique has been proposed to show better robustness against rotation attacks and other issues in watermarking. In this work, the Deep Belief Network (DBN) is trained by the Bear Smell Search Algorithm (BSSA) during the first phase. The second phase is the embedding phase, which utilizes the hybrid transform domain of Discrete Wavelet Transform and Singular Value Decomposition (DWT-SVD). After embedding, the extraction is performed with the help of a Back Propagation Neural Network (BPNN). The experiments are conducted on four kinds of host images such as Barbara, Peppers, Person and Lena. Experimental results proved that the superiority of our proposed method achieving the state-of-the-art accuracy.

Combined image encryption and steganography technique for enhanced security using multiple chaotic maps

This paper proposes a novel security algorithm which combines the concepts of Cryptography and Steganography to provide enhanced security to digital images. The proposed algorithm, named CIEST, enciphers the secret image (cryptography) which is hidden into the cover image (steganography) for secured transfer of information. A two-stage encryption process, with one of the stages being DNA encoding is used here. The cover image is also modified before the data embedding phase to improve the effectiveness of the algorithm. Multiple Combined Chaotic maps are utilised in the encryption and embedding stages to improve the randomness of the algorithm. The use of chaotic maps also makes the algorithm more robust to different types of attacks. The CIEST scheme is shown to be robust, offering high embedding capacity with high sensitivity with a near zero correlation value, information entropy value of 7.9972, maximum SSIM of 0.9992, and a maximum PSNR of 56.1071. The experimental results presented establish the immunity of the proposed method against attacks along with the feasibility and effectiveness of the algorithm in real time.

A fusion recommendation model based on mutual information and attention learning in heterogeneous social networks

Heterogeneous social networks generate vast amounts of data due to their large user base. In response, recommendation systems have been developed as an effective tool to mitigate the problem of information overload. Nonetheless, standalone recommendation methods are often limited by their own inherent deficiencies. Consequently, deep learning has emerged as a powerful approach to address these issues. By integrating diverse recommendation methods, the shortcomings of individual methods can be partially compensated for and improved recommendation outcomes can be achieved. This paper presents a novel user interest fusion recommendation model called MIAL, which leverages mutual information and attention-based learning mechanisms. Specifically, a user’s posts are first transformed into word sequences via a Convolution Neural Network (CNN) based interest recommendation model for pre-processing. To enhance the multi-label prediction model, this study adopts a label selection method based on the label maximum mutual information spanning tree (LMMIST) to determine the final recommended label set. To identify reliable user similarity, the concept of pseudo-similar friends is introduced to construct heterogeneous networks. During the network embedding phase, a new random walk algorithm termed MRWalk and a pseudo-similar propagation mechanism are proposed. Experimental results using the Twitter dataset and public Amazon product dataset demonstrate that the proposed MIAL algorithm outperforms various state-of-the-art algorithms. The comparative analysis shows that MIAL has significantly improved performance.

Molecular insight into toughening induced by core-shell structure formation in starch-blended bioplastic composites

Binary and ternary blends with poly(lactic acid) (PLA), poly(butylene succinate) (PBS), and thermoplastic starch (TPS) were prepared by a melt process to produce biodegradable biomass plastics with both economical and good mechanical properties. The mechanical and structural properties of each blend were evaluated. Molecular dynamics (MD) simulations were also conducted to examine the mechanisms underlying the mechanical and structural properties. PLA/PBS/TPS blends showed improved mechanical properties compared with PLA/TPS blends. The PLA/PBS/TPS blends with a TPS ratio of 25–40 wt% showed higher impact strength than PLA/PBS blends. Morphology observations showed that in the PLA/PBS/TPS blends, a structure similar to that of core-shell particles with TPS as the embedding phase and PBS as the coating phase was formed, and that the trends in morphology and impact strength changes were consistent. The MD simulations suggested that PBS and TPS tightly adhered to each other in a stable structure at a specific intermolecular distance. From these results, it is clear that PLA/PBS/TPS blends are toughened by the formation of a core-shell structure in which the TPS core and the PBS shell adhered well together and stress concentration and energy absorption occurred in the vicinity of the core-shell structure.

Abstractive Summarization of Text Document in Malayalam Language: Enhancing Attention Model Using POS Tagging Feature

Over the past few years, researchers are showing huge interest in sentiment analysis and summarization of documents. The primary reason being that huge volumes of information are available in textual format, and this data has proven helpful for real-world applications and challenges. The sentiment analysis of a document will help the user comprehend the content’s emotional intent. Abstractive summarization algorithms generate a condensed version of the text, which can then be used to determine the emotion represented in the text using sentiment analysis. Recent research in abstractive summarization concentrates on neural network-based models, rather than conjunctions-based approaches, which might improve the overall efficiency. Neural network models like attention mechanism are tried out to handle complex works with promising results. The proposed work aims to present a novel framework that incorporates the part of speech tagging feature to the word embedding layer, which is then used as the input to the attention mechanism. With POS feature being part of the input layer, this framework is capable of dealing with words containing contextual and morphological information. The relevance of POS tagging here is due to its strong reliance on the language’s syntactic, contextual, and morphological information. The three main elements in the work are pre-processing, POS tagging feature in the embedding phase, and the incorporation of it into the attention mechanism. The word embedding provides the semantic concept about the word, while the POS tags give an idea about how significant the words are in the context of the content, which corresponds to the syntactic information. The proposed work was carried out in Malayalam, one of the prominent Indian languages. A widely used and accepted dataset from the English language was translated to Malayalam for conducting the experiments. The proposed framework gives a ROUGE score of 28, which outperformed the baseline models.

Increase Security of Image Encryption Depends on Change Pixel Location

In any communication, security is the most important issue in today’s world. The information security has be come one of the most signify can tproblemsin data communication. Lots of datasecurity and data hiding algorithmshavebeendevelopedinthelastdecade. Cryptography and steganography are the two major techniques for secret communication. In this paper, the secret text is first encryptedbyusingAESwhichhasverygood performanceandisamostpowerfultechnique. Now this encrypted text is embedded using steganography. Our proposed system gives threestages of security for secret. The main aim of proposed method to increase security of embedding and extraction phase using AESencryption,steganography and change pixel location. In order to evaluate performance the proposed algorithm performs series of tests. These tests include, PSNR and MSE.