Cloud Computing and Security

Feature extraction of ciphertext is a key procedure in cryptosystem recognition task. Varieties of ciphertext’s features are proposed in exited literatures, while feature based on randomness test has derived little attention. In this paper, by segmenting ciphertexts and changing parameter of randomness test, we propose 54 features of ciphertext based on NIST’s 15 randomness tests. As a measure of these features, we choose support vector machine as classifier algorithm to verify its performance in cryptosystem recognition. In experimental settings, we consider 15 situations of 6 cryptosystems’ one to one recognition. The experimental results demonstrate that the application of randomness test in cryptosystem recognition is feasible and necessary. Most of proposed features reach better recognition accuracies than random recognition, which indicates that randomness tests are applicable for cryptosystem recognition applications. And we also conduct further analysis: (a) analyze some features’ recognition performance, and find the relation of some feature’s recognition accuracies and cryptosystem. (b) compared with existed features, part of new features maintain high recognition accuracy with lower dimension and smaller data storage space.

Since the proxy re-encryption has the limitation of distributed applications and the security risk of collusion attacks in semi-trusted distributed environments (e.g. cloud computing), the novel definition of threshold proxy re-encryption is proposed based on secret sharing and proxy re-encryption. According to the definition, the threshold proxy re-encryption scheme can be flexibly created with the standard cryptographic prototype. An efficient, secure, and implementable unidirectional threshold proxy re-encryption scheme is constructed by the combination of Shamir’s secret sharing, and is proved secure by using the intractability of discrete logarithms. This paper presents a consortium blockchain access permission scheme, which is built on the threshold proxy re-encryption scheme. When a new node joins a consortium blockchain, an access permission is achieved by the agreement on other existing nodes, instead of a centralized CA.

The discrete Fourier transform and trace representation of certain sequences can help generate the sequences efficiently and analyse their cryptographic properties. In this paper, we first determine the defining pairs of the binary sequences derived from a class of polynomial quotient modulo an odd prime p and the Legendre symbol. We then derive the discrete Fourier transform and the trace representation of this class of sequences.

Various results of the no-cloning theorem and no-superposing theorem in quantum mechanics have been proved using the superposition principle and the linearity of quantum operations. In this paper, we investigate general transformations forbidden by quantum mechanics in order to unify these theorems. We prove that any useful information cannot be created from an unknown pure state which is randomly chosen from a Hilbert space according to the Harr measure. And then, we propose a unified no-go theorem based on a generalized no-superposing result. The new theorem includes various no-go theorems of the no-cloning theorem, no-anticloning theorem, no-splitting theorem as a special case.

We are living in the age that crisis events happen everywhere and every day. Video evidence plays an important role in restoring the truth of the incident. The credibility of video evidence is gradually declining. On the one hand, the credibility of the government has been questioned. On the other hand, malicious attackers will distort the video and publish it on the Internet to mislead public. Inspired by the blockchain, this paper proposes a model based on blockchain to ensure the credibility of video evidence, which is called VideoChain. VideoChain is essentially a blockchain system, but different from digital currency such as Bitcoin. It records the hash of the surveillance video. This paper takes the campus Video Surveillance as an example to describe the architecture and operation of VideoChain. We also give a consensus protocol based on the proof of stake according to the application scenario. The analysis shows that VideoChain is traceable and non-tampering, which can effectively ensure the credibility of video surveillance data. It also has good operational efficiency. This research is of great significance to enhance the trust between the government and citizens, which will improve the credibility of the government.

With the development of the big data age, information security is becoming more and more important. Screen content image are composed of text, graphics and natural image. They present strong anisotropic features, especially on the text and graphics parts. It is well known that Spread Transform Dither Modulation(STDM) is more robust to re-quantization, such as JPEG compression, than regular Quantization Index Modulation(QIM). In this paper, we propose a novel watermarking scheme for screen content grayscale image in DCT domain. On the basis of STDM, combined with the characteristics of human visual system, we use the texture complexity effect factor on DCT domain to adjust the watermarking process. To evaluate the performance of our proposed scheme, we use the reference image from the SIQAD image database. The 20 reference SCIs were thoughtfully identified from the Internet, and they cover a wide variety of image contents, including texts, graphics, symbols, patterns, and natural images. Experiments show that our method has a good performance in term of robustness and better visual quality.

As the main active traffic analysis method, network flow watermarking (NFW) has been proven effective for flow correlation in anonymous communication system or stepping stone detection. In various types of network flow watermarking schemes, the interval-based ones can achieve significant better capability of resisting network interference. However, there still exists no work to give a comprehensive analysis of them, specifically on practicability as the implementation of NFW in Internet still remains a great challenge. In this paper, the existing interval-based NFW schemes are comparatively analyzed by benchmarking their performance on robustness, invisibility and practicability. Different from some prior work, we pay special attention to the practicability evaluation, which is related to time and storage overhead, communication and computation overhead, and the statistical model demand. Experimental results on CAIDA dataset give an overview of the existing interval-based NFW schemes.

Covert communication using Internet Protocol version 6 (IPv6) header fields can be easily detected. By thoroughly exploring the characteristics of IPv6 multicast, this study proposes a novel covert communication model based on IPv6 multicast (MCv6). In this model, a multicast group, containing a large number of members across different subnets, is created to hide the receiver’s network ID, thereby achieving covert communications. To ensure the security of this covert communication, a random key generation algorithm, based on the chaotic sequence, is proposed to encrypt communication packets. To ensure the legitimacy of covert communications, a multicast source authentication mechanism based on hash comparison is proposed to verify the legitimacy of communication source nodes. To ensure the integrity of covert communications, a two-stage error control mechanism is proposed to control the possible packet-loss and other errors. Theoretical analysis and simulation results show that the proposed MCv6 model can provide good IPv6-based covert communications, efficiently reducing the probability of detection, and ensuring the security and reliability of the IPv6-based medium.

Visual quality and bit-rate increase are two main metrics of evaluation for many researchers to evaluate the marked videos that additional data is embedded into. In this paper, we combine matrix embedding with several assumptions to propose a novel data hiding method in H.264/AVC video stream. The assumptions can be exploited to analyse the propagation of intra-frame distortion. Namely, both matrix embedding and the given assumptions are used for improving the quality of marked videos. Therefore, although the proposed scheme cannot completely avoid the intra-frame distortion drift, it can keep a few degradations in terms of visual quality and keep a small variation in bit-rate increase. Experimental results have verified that the proposed method has no significant degradation in terms of visual quality (i.e., PSNR and SSIM) and indeed obtains a very small variation in bit-rate increase.

Tor hidden services have grown too rapidly so that anonymous users cannot verify the authenticity of Hidden Services. Meanwhile, there are many unsafe factors in the centralized directory management mode, and the service content and configuration can be detected and monitored by automated scripts. Faced with the above problems, this paper based on the current Tor domain name communication management model, proposed a new Hidden Service Domain Name System—HSDNS, which using SOR node decentralized management, then replace the top-level domain name as “.hs” to increase the anti-scanning property of the original domain name, introduces the PoW competition mechanism to enable registrants to get readable unique domain name among the whole network. On this basis, the random number provided by the BTC blockchain guarantees random competition and enhances the anti-registration attack, in the meantime, Merkle Tree hash detection mechanism is used to open verification service of domain authenticity, and finally verify model analysis results through the ExperimenTor simulation platform, and it shows that HSDNS model has decentralized management, service anti-scanning, domain name uniqueness and authenticity verification and so on.

The aim of steganography is to put a secret into carriers and only be seen by participants. This paper proposes a layered steganography model which helps to simplify the steganography design. The steganography model enables the two communication parties to interact with each other, which enables the receiving end to get data by data prediction, and optimizes the steganography mechanism to take full advantage of the existing methods to implement multi-carriers steganography. Herein, the interactive steganography action is decomposed into three basic interactive procedures, and the embedding and extracting procedures are implemented via these basic interactions, in which texts, pictures, voices or videos are used as carriers. With the features of language diversification, common media, real-time in an instant message communication, the proposed steganography in this paper is flexible, secure and reliable.

In this paper, a high-performance error-prediction method based on Multiple Linear Regression (MLR) algorithm is first proposed to improve the performance of Reversible Data Hiding (RDH). The MLR matrix function that indicates the inner correlations between the pixels and its neighbors is established adaptively according to the consistency of pixels in local area of a natural image, and thus the object pixel is predicted accurately with the achieved MLR function that satisfies the consistency of the neighboring pixels. Compared with conventional methods that only predict the object pixel with fixed parameters predictors through simple arithmetic combination of its surroundings pixel, experimental results show that the proposed method can provide a sparser prediction-error image for data embedding, and thus improves the performance of RDH more effectively than those state-of-the-art error prediction algorithms.

The content of the document tends to be tampered and forged easily, and it may cause serious consequences in many important situations. Therefore, this paper proposes a multiple watermarking scheme for copyright protection and tamper detection of OOXML format documents. For an OOXML format document, we firstly embed a robust watermark by synonym substitutions, then embed double watermarks fragilely by utilizing the format characteristics of the document. The double fragile watermarks are embedded by the ways of modifying the word number in a text element defined by the main document body and modifying the values of revision identifiers. The embedding of multiple watermarks does not cause any visual impact on the content and format of the original document, which maintains the original meaning of the content basically. As shown by the experimental results, the watermarking method based on synonym substitutions has a strong robustness against the attacks on formats and the content. However, the double fragile watermarking methods can sensitively perceive tampers in the content and formats of the watermarked document, and predict the type of tampers with capability of locating the tampering.

This paper is to investigate a novel framework of robust video watermarking based on the statistical model with robustness against multiple attacks. The main contribution is threefold. First, the Laplacian distribution is proposed to model each naive video frame, referring to as the original frame; meanwhile the noisy frame, referring to as the one with adding Gaussian-distributed noise, is modeled using the Gaussian distribution. Second, we propose a novel mechanism of embedding watermark by artificially adding noise or not, corresponding to watermark bit 1 or 0. Third, it is proposed to cast the problem of watermark extraction into the framework of hypothesis testing theory. In the ideal context, with knowing all the model parameters, the Likelihood Ratio Test (LRT) is smoothly established with verifying the feasibility of the designed watermark extraction based on the statistical models. In the case of estimating model parameters, we propose to design the Generalized Likelihood Ratio Test (GLRT) to deal with the practical problem of watermark extraction. Finally, compared with some prior arts, extensive experimental results show that our proposed novel framework of robust video watermarking can achieve the high video quality with robustness against various attacks such as re-scaling, cropping, and compression.

In textile industry production, fabric defect inspection is a vital step to ensure the quality of fabric before spreading, cutting and so on. Recently, image characteristic of nonlocal self-similarity (NSS) is widely applied to image denoising due to its effectiveness. Actually, fabric defect detection can be considered as a problem that finds noises in an image. Based on the reason, we propose a simple yet effective method, namely nonlocal low rank approximation (NLRA), for fabric defect detection. In NLRA, an image to be processed is divided into many patches. For a given patch, we search its several similar patches and group them as a matrix. Then, the clean image patch can be reconstructed through solving the low rank approximation of the matrix. Finally, a new image will be synthesized from these estimated patches, the defects can be located by finding the difference between the original fabric image and the reconstructed image. Experimental results prove the validity and feasibility of the proposed NLRA algorithm.

Recently, many high security steganographic schemes have been developed to approximate non-additive distortions with consideration of mutual adjacent modification influences. However, their security performance behaves volatilely and is very sensitive to the heuristically designed parameters and the initial additive distortion function. To make a stable security performance and to further improve the embedding security, a novel model-based steganographic scheme is proposed by incorporating the adjacent embedding information with no dependence on the heuristic parameters. We first divide the cover image into several interleaved sub-lattices, and then optimally embed message into each sub-lattice in sequence. In each sub-lattice, the embedding change rates for each pixel are optimized by utilizing the adjacent modifications as priori knowledge. During the optimization process, a novel asymmetric probability model is designed to simultaneously tackle with two inequivalent change rates for modifying the cover pixel by $$+1$$ and $$-1$$ . Experimental results show that the proposed scheme can rival or outperform the prior arts, and moreover, can provide a stable security performance.

The widespread distribution of multimedia data cause copyright problems for digital content. This study makes use of digital image watermarking technology to protect copyright information, and proposes a scheme utilizes the support vector machine (SVM) based on posterior probability and the optimized genetic algorithm (GA). Firstly, each training image is divided into sub-blocks of 8 * 8 pixels, and they are trained and classified by the SVM to obtain the adaptive embedding strength. Secondly, after the operation of reproduction, crossover, mutation, the genetic algorithm generates new individuals in the search space by selection and recombination operators to optimize the objective function, and find out the best embedding position of the watermark. The 8 * 8 pixel sub-blocks were transformed by DCT when embedding. Finally, the watermark is extracted according to the embedding rules. Compared with the experimental results of other algorithms, the proposed scheme has better resistance against some common attacks, such as Histogram Equalization, Guassian Noise (0.04), Guassian Noise (0.05), JPEG (QF = 50), Salt-pepper Noise (0.01).

Digital watermarking is considered to be a potential and effective solution to protect digital content. The key issue of a watermarking system is the trade-off between the robustness and transparency. Therefore, in order to improve these two characters, this paper proposes a digital image watermarking scheme based on discrete cosine transform (DCT) and discrete wavelet transform (DWT). After the host image was transformed by DWT, the low frequency band was selected and divided into 8 × 8 blocks. And then, each block was transformed by DCT. Meanwhile, the watermark was improved security by being encrypted using ergodic matrix, thus it cannot be recognized after the decryption. At the embedding process, the DCT coefficients at the five rows of the bottom of each block were exchanged by comparing the scrambled watermark information. At the extracting process, the watermark was recovered by comparing the value of two adjacent coefficients in the DCT block. Experimental results show that the proposed scheme has good imperceptibility as well as good robustness against common watermark attacks, such as Gaussian low-pass filtering, histogram equalization, image brighten/darken, contrast decreasing, salt-pepper noise, average filtering and cropping.

Digital watermarking technology in the field of information hiding technology has become an important means to protect copyright in network transmission. Robust reversible digital watermarking has been studied for many years as a more adaptive technique for network lossy transmission environment. However, the algorithms proposed by the scholars have not done a good job in balancing the embedded distortion and robustness of the human eye. In this paper, an adaptive robust reversible digital watermarking technique is proposed, which distinguishes the texture complex region and the texture smoothing region by the method of complexity prediction, and selects different parameters to realize the function of embedding more bits in the complex texture region and embedding less bits in the smooth region. In this paper, the algorithm has a high robustness to better resist JPEG compression while there is less distortion of human eye observation. Experimental results show that the proposed scheme has better robustness and higher subjective quality than previous schemes.

In this paper, a reversible data hiding scheme for digital images with high hiding capacity is proposed. Original image is segmented into smooth and rough regions based on local complexity. In order to achieve higher hiding capacity, we embed three bits into each pixel belonging to smooth region with lower local complexity and one bit is embedded into each pixel of rough region, which can effectively exploit more redundancy during data embedding compared with conventional methods of prediction error expansion (PEE). Additionally, the pixel selection mechanism is applied to reduce the number of shifted pixels, which leads to high visual quality of stego image. Experimental results show that, our scheme can achieve better rate-distortion performance than some of state-of-the-art schemes.

Content authentication of the text document has become a major concern in the current digital era. In this paper, a tamper locating algorithm for DOCX document content authentication is proposed. Firstly, according to the characteristics of DOCX format, the authentication information unrelated to the text content is embedded into the main setting file named document.xml by displaying characters segmentation. Then, identify the integrity of the text by confirming whether the embedded watermark is same to the authentication watermark. Experiments show that the algorithm is very fragile to any modification and can locate the tampered places very well.

Video information hiding has attracted increasing attention in the field of information security. In the literature, great quantities of information hiding algorithms based on DCT and intra prediction mode have been proposed for the latest video coding standard - High Efficiency Video Coding (HEVC). However, few algorithms are reported to hide information in the inter prediction mode, one of the unique advantages of HEVC. This paper proposes an information hiding algorithm based on PU partitioning mode in P-frames, which realizes the embedding of hidden information without affecting the video quality. The PU partitioning mode selected by HEVC based on the optimum CU structure is recorded in the first round calculation and adopted as the reference for the modification process. PU partitioning modes are altered to embed hidden information. Since only the PU partitioning modes of P-frames is modified, the proposed algorithm can be combined with existing DCT-based or intra-prediction-based algorithms to greatly increase the embedding capacity while keeping the quality of the video sequence not being affected.

Deep neural networks (DNN) have achieved state-of-art performance on image classification and pattern recognition in recent years, and also show its power on steganalysis field. But research revealed that the DNN can be easily fooled by adversarial examples generated by adding perturbation to input. Deep steganalysis neural networks have the same potential threat as well. In this paper we discuss and analysis two different attack methods and apply the methods in attacking on deep steganalysis neural networks. We defined the model and propose the concrete attack steps, the result shows that the two methods have 96.02% and 90.25% success ratio separately on the target DNN. Thus, the adversarial example attack is valid for deep steganalysis neural networks.

Recent studies have shown that the internal composition of the Chinese word provides rich semantic information for Chinese word representation. The Chinese word consists of one or more Chinese characters. Chinese characters have semantic information. And some Chinese characters have multiple meanings. Moreover, the composition of Chinese characters has different semantic contributions to word. In response to this phenomenon, this paper proposes a new attention-based model (ACWE) to learn Chinese word representation. At the same time, the “HIT IR-Lab Tongyici Cilin (Extended Version)” can calculate the semantic similarity between Chinese characters and words. And it can reduce the impact of data sparseness and improve the effectiveness of Chinese word representation. We evaluate the ACWE model from the similarity task and the analogical reasoning task, and the experimental results show that the ACWE model is superior to the existing baseline model.

Most state-of-the-art text steganography algorithms are designed based on synonym substitution with the concern of simplicity and robustness. However, synonym substitution will cause some detectable impact on cover texts. In this paper, we propose an content-adaptive text steganography to minimize the impact caused by embedding process. We believe that synonym substitution will cause a hybird distortion consists of statistical distortion and linguistical distortion. We design a double-layered STC embedding algorithm (HSL) to minimize the distortion. Experiments results indicate that the security performance of HSL is better compared with traditional methods based on synonym substitution.

In textile industry production, fabric defect inspection is a very important step to ensure the quality of fabric. At present, most of the methods can detect the defects for solid color with the distinguishable defects, but they are not very efficient for small defects, especially for the defects which has small difference with the background. In this paper, we propose a three-phase method, mean filter, convolution operator combined with variance (MCV), for fabric defect detection. For a fabric image, we first use mean filter to suppress noise, then convolution operator is applied to enhance image. Based on enhanced image, we divide it into many patches. For a given patch, we calculate its variance and then use the threshoding to decide whether the patch is free defect or not. Finally, a defect image will be synthesized from these processed patches. Experimental results prove the effectiveness of the proposed MCV algorithm.

Recent studies have proved a well-designed convolutional neural network (CNN) is a good steganalytic tool. In this paper, based on the previous work, we report a method using segmented subimages to improve the testing accuracy of CNN for steganalysis. In training phase, a CNN is trained on training set of whole image. In testing phase, for a given testing image, a sliding window is employed to segment the whole testing image into subimages. Each subimage is feed into the trained CNN respectively to obtain a subdecision. The final decision is obtained through majority vote. Experiments show that the proposed method achieves significant improvement on testing accuracy when detecting S-UNIWARD and HILL under payload of 0.4 bpp, whereas the time efficiency is only slightly worse compared with previous work.

Protocol reverse engineering is of great significance for discovering protocol vulnerabilities, improving protocol security and reusing protocol. The existing protocol reverse analysis methods usually need a great deal of computation and often takes a long time, which seriously affects the effect of real-time analysis. This paper proposes an incremental protocol format extraction algorithm, which divides the network traffic into different substreams, and introduces error decision mechanism to avoid local errors caused by partition, so as to ensure the correctness. By dynamic evaluation of the complexity of the protocol analysis, the incremental protocol analysis method can effectively improve the efficiency of the protocol reverse engineering.

Medical image containing patient information is often faced with various attacks in the transmission process. In order to enhance the medical information system security, and effectively solve the problem of medical data protection, a new algorithm of medical image watermarking based on SIFT-DCT perceptual hashing (scale invariant feature transform and discrete cosine transform) is proposed. Firstly, use SIFT-DCT perceptual hashing to extract features for the original medical images and quantize to generate hashing sequences. Then, use chaotic maps to encrypt the watermarking and embed it in the medical image. Finally, calculate the correlation coefficients of the embedded and extracted watermarking sequences to reflect the robustness of the algorithm. The results of experiment show that the proposed algorithm has good robustness against conventional attacks and geometric attacks, especially in terms of rotation, translation and clipping.

Aiming at the problem that the existing network storage covert channel detection algorithm can not take into account both the detection rate and the computational complexity, a network storage covert channel detection method based on data joint analysis is proposed. This method studies the information hiding mechanism of the network storage covert channel according to related documents. Based on this, the regularity characteristics of the packets in each field of the network data packet and the correlation characteristics between the packets are analyzed. The above characteristics are further transformed into eigenvector matrices through kernel density estimation, variation coefficient, fragility entropy, and autocorrelation coefficient. And SVM classifier is trained using eigenvector matrices. The experimental test shows that this method has a high detection rate and its computational complexity is small.

In order to improve the effective utilization of available resources in the traditional wireless network, this paper studies the optimization of resource allocation (RA) in the underlay cognitive radio network (CRN). Our goal is to maximize the sum rate of the whole system (e.i., primary users (PUs) and secondary users (SUs)), taking into account the constraints of interference temperature (IT) and minimum rate, and the Quality of Service (QoS) guarantees. A heuristic algorithm is proposed to solve the non-convex non-linear programming optimization problem. Theoretical analysis and simulation results show that this algorithm can effectively reduce the power interference to the PUs, maximize the transmission rate of PUs and SUs, and improve resource utilization of the CRN.

With the continuous development of Internet technology, a joint method of data embedding and scrambling has become a new research trend. As the most frequently used image in our daily life, JPEG is applied in many ways. In this paper, we present a new reversible data embedding and scrambling scheme based on JPEG images. We use the histogram shifting method to embed the secret data for the AC coefficients after decoding the JPEG image. And for the DC coefficients, we apply the re-encoding method to embed the data, and then scramble the DCT blocks of the JPEG image. The experimental results show that we can easily achieve high embedding capacity and scrambling effect by using the proposed scheme. At the same time, the original image can be perfectly reconstructed after the secret data extraction.

Difference expansion (DE) and histogram modification (HM) are efficient ways for reversible data hiding (RDH) into digital video. In most occasions, the reversibility of an algorithm cannot be confirmed until the data extracting experiments have been carried out, it means the design of a reversible data hiding algorithm lacks theoretical guidance. In this paper, by studying some typical algorithms, we presented a method called shifting mode diagram. From the shifting mode diagram, we can judge the reversibility of an algorithm by three rules, and more importantly, one can design a reversible data hiding algorithm or optimize the existing methods by using the rules. According to the characters of shifting mode diagram, we summarized a number of formulas, with the formulas, we can estimate the embedding capacity with a high accuracy, experiments have proven that, and the factors that influence the accuracy are also studied.

The joint photographic experts group (JPEG) is the most popular image format in our daily life, and it is widely used by digital cameras and other photographic capture devices. Recently, reversible data hiding (RDH) for JPEG images has become an active research area in the field of data hiding. In this paper, a new two-dimensional coefficient-histogram based RDH scheme for JPEG image is proposed. First, a two-dimensional quantized discrete cosine transform (DCT) coefficient-histogram is generated. Then, data are embedded according to a specifically designed coefficient-pair-mapping (CPM). Here, by the proposed approach, compared with the one-dimensional histogram-based RDH for JPEG images, the increased file size is minimized. Moreover, the selection strategy based on the optimal frequency band of the DCT coefficient-pairs is proposed, by which the distortion of the marked JPEG image is minimized. Compared to some state-of-the-art RDH methods for JPEG images, experimental results show the superiority of our methods both in image quality and increased file size.

In reversible data embedding, to avoid overflow and underflow problem, before data embedding, boundary pixels are recorded as side information, which may be losslessly compressed. The existing algorithms often assume that a natural image has few boundary pixels so that the size of side information could be rather small. Accordingly, a relatively high pure payload could be achieved. However, there actually may exist a lot of boundary pixels in a natural image, implying that, the size of side information could be very large. Thus, when to directly use the existing algorithms, the pure embedding capacity may be not sufficient. In order to address this important problem, in this paper, we present a new and efficient framework to reversible data embedding in images that have lots of boundary pixels. The core idea is to losslessly preprocess boundary pixels so that it can significantly reduce the side information. We conduct extensive experiments to show the superiority and applicability of our work.

A robust H.264/AVC video watermarking algorithm without intra distortion drift is proposed in this paper for the copyright protection of digital videos. The classic distortion drift-free method limits the size of watermark image because of limited capacity. The improved intra distortion drift free method proposed in this paper enlarges the capacity and promote the visual quality with a reasonable classification according to intra prediction modes of H.264. Embedding pretreated watermarks into middle-frequency coefficient-pairs of $$4\,\times \,4$$ luminance blocks, which makes the scheme achieve robustness. Experimental results show that our algorithm achieve high robustness and good visual quality without huge bit-rate increasing. The capacity of proposed scheme is doubled than the classic scheme as well.

This paper proposes a novel steganographic method to hide secret data during the generation of marbling patterns. We select some points on white paper to represent secret information. These points are connected with lines to construct an original pattern. With a series of deformation operations, the original pattern is transformed to generate a marbling pattern. During the process of data hiding, we construct a unit library containing different deforming operations. The library records the parameters of each deformation, and defines the mapping between the binary data and the deformation type. The unit library is shared with the recipient so that the deformation parameters can be recovered correctly. After using reverse deformations, the recipient identifies the location of the inflection points and extracts the secret data. Experimental results show that the proposed method performs high security and flexible embedding capacity. Meanwhile, the marbling images have a good visual effect. Furthermore, the proposed steganography provides a capability of countering JPEG compression.

This paper proposes a joint scheme of data hiding and style transfer, which embeds secret data during the procedure of transferring natural images into comic styles. While most data hiding algorithms employ clean images as covers, we employ the processed images that are popular on social networks. The style transfer based data hiding includes two phases. In the first phase, we brighten the image and remove the details by enhancing the saturation and smoothing the content. In the second phase, we propose an edge marker embedding based algorithm to enhance the contours and generate comic-style stego images. Experimental results show that the proposed approach provides a large embedding capacity and a good capability of resisting steganalysis.

A barrel shifter is a common component of high-speed processor, which can realize the displacement operation of the specified number of data word in a single cycle. On the basis of the inverse logic circuit, a displacement device with n inputs and m control bits is proposed, which is denoted as (n, m) shifter, and a set of control inputs that specify how to shift in data between input and output. On the basis of the quantum reversible logic circuits, for synthesizing the barrel shifter, we present the novel method based on the decomposition of the permutation group and some Construction Rules. It only uses (3, 1) shifter and controlled swap gate to quickly synthesize any controlled shifter with low quantum cost, and any (n, k) barrel shifter can be got by cascading the least of k corresponding (n, 1) shifters. The quantum circuit shifters generated by this method can reduce the number of quantum gates, reduce the quantum cost and improve the efficiency of the algorithm, so that all kinds of reversible barrel shifter can be rapidly designed. In this article, we mainly give the ways on qubit left circular shifts, bit permutation and line permutations, and other types of basic shift circuits are also designed.

Coverless information hiding does not make any modifications to the carrier, so it can effectively resist the various steganalysis and detection algorithms. However, the existing methods still have some problems that include low hiding capacity and unsatisfactory hiding success rate. To address these problems, this paper proposes a method of text coverless information hiding based on word2vec. The method uses distance algorithm provided by word2vec to obtain similar keywords, then utilize similar keywords to enlarge the set of keywords, finally retrieve the stego-texts that contains the combination of the location tags and the keywords. The experimental results and analysis show that the method can ensure the hiding success rate of 100%, and the hiding capacity is 2.87.

Open Office Xml (OOX) is the Microsoft Office document format type, which has been widely used in the world. It is of great significance to research the OOX document information hiding technology. In this paper, a text information hiding algorithm was proposed. The proposed method used OOX document custom components to hide secret information. Because the custom component was not referenced by the main document, the content and format of the main document have not been modified. In addition, since the information hiding process does not cause any change in the content and semantics of the displayed text in the document, it can resist content-based attacks, semantics-based attacks, and format-based attacks.

Most state-of-the-art detection methods against synonym substitution based steganography extract features based on statistical distortion. However, synonym substitution will cause not only statistical distortion but also semantic distortion. In this paper, we propose word embedding feature (WEF) to detect the semantic distortion. Furthermore, a fused feature called word embedding and statistical feature set (WESF) which consists of WEF and statistical feature based on word frequency is designed to improve detection performance. Experiments show that WESF can achieve lower detection error rates compared with prmethods.

At present, the research on the evaluation methods of communication jamming effectiveness is in its infancy, a comprehensive evaluation method based on Grey relational analysis method is proposed in this paper. Firstly, signal characteristics of the jamming signals are analyzed and evaluation indexes are proposed. Then, this paper introduces the Grey relational analysis and the data processing method. In the end, evaluation process of jamming schemes is proposed and the simulation verifies the rationality and feasibility of this evaluation method.

The frequency dependence of the antenna voltage patterns and the amplitude and phase errors calibrated by external noise sources is analyzed and verified by simulations and experiments, an improved calibrated method using external single signal is proposed to be an alternative to the external noise sources, and the simulations and experiments show that the improved method can mitigate the effect of the frequency dependence of that can be as an alternative in the calibration of aperture synthesis radiometer.

In recent years, with the rapid development of agricultural Internet of things, many different agricultural and sideline products can be bought easily which from different regions, but it also brings about more problems of foods safety (production and sale of fake eggs, water injection Pork, etc.). These issues affect social development and people’s health seriously. According these issues, all stages of agricultural products should be monitor and control. This paper introduces the traceability system which integrates RFID and EPC technologies, and uses ONS database to store key information. The tracing information can be queried that user login in traceability system and identify the two-dimensional code. People use the mobile scan to encrypted the QR code to get the information of agriculture and food products they need. The traceability system guarantees the safety of meat in all aspects of the supply chain. The system should be developed and promoted in society.

Biometric technology is an important characteristic and a reliable authentication method, which can be used to verify the user’s identity for its accessibility and uniqueness. In this paper, we analyze the existing protocols and find that they still have some security flaws. In order to effectively enhance the security, we present a biometrics-based remote user authentication scheme using smart cards to overcome those weaknesses. According to security and cost analysis, compared with other authentication schemes, the proposed scheme costs $$21T_h$$ . However, it can implement more security goals and withstand different attacks, such as DoS attack, anonymity attack and leak of verifier attack.

In this paper, a novel backscatter link frequency (BLF) generation scheme is presented. The accuracy of BLF required by EPC Class-1 Generation-2 (Gen2) is one of the critical issues in UHF RFID tag design. By analyzing the effects of division ratio and division error on the accuracy of BLF, a novel BLF generation scheme is proposed to reduce the BLF errors. Simulation results show that the BLF generated by the proposed scheme can satisfy the requirement of EPC Gen2 standard when the clock frequency is no less than 1.632 MHz, which significantly simplifies the design complexity of clock generator.

This paper introduces a platform development and implementation method which is based on block chain to protect mobile phone privacy. It is applied to collective maintenance, programmable, safety and reliability, and the requirements of personal privacy protection. The platform consists of the manufacturer node, market node, mobile phone node and user node, which is connected by a star topology construction. All nodes include data collection module, data storage module, target detection module and data application module, which is matched by the publish/subscribe mode. Each node will match a distributed storage database for storing various transaction data. The data storage module realizes the communication between nodes and completes the building of mobile phone protection platform. This method can solve the problems such as get the usage right by flashing root, user information leakage and so on. Finally, we give the application scenario of mobile phone protection platform, aiming at realizing the privacy protection on the block chain.

As one of the fundamental operations, matrix multiplication plays a significant role in mathematics, computer science and many other science fields. In Williams’ research of studying matrix multiplication problem, she put emphasis on studying the even tensor powers of Coppersmith-Winograd approach, and then obtained improved upper bound for the matrix multiplication exponent. In fact, the program for calculating the so-called even tensor power is a constrained optimization problem with complicated constraints. In this paper, we focus on the 4-th tensor power problem of matrix multiplication. After converting this practical problem, we design a dominance-based constrained optimization evolutionary algorithm. Empirical results show that this algorithm can effectively solve the 4-th tensor power problem. What is more, the feasible solution obtained by this algorithm is better than the current known solution of the problem.

The rapid development of agricultural Internet of things is difficult to deal with a lot of information. This paper proposes a method based on ontology of agricultural network information fusion, from this point the basis of ontology and information fusion classification and methods, basic principle, technology in agricultural Internet information fusion as a foundation, in view of the agricultural Internet information uncertainty, heterogeneity and representation problem, put forward agricultural content networking information fusion method based on ontology. It provides theoretical support for the information processing of agricultural Internet of things.

In classic operation systems, processes are assigned different privileges according to the resources. The enforcement of privilege differentiation on diverse processes indicates that strict security management on the individual process, whose emphasis on the restriction on respective process, however, may also overlook the security risk among the processes. Specifically, one process can invoke another one and establish a session, during which the privileges of invoked process may be passed to the invoking process (e.g., by the inter-processes requests). Thus, it may result in the abuse of privilege and resource leakage. Moreover, the inter-networking of the processes and their relations also complicate the tasks for the regulation on authorized privileges, and those can be obtained by inheritance. The management on the latter case (i.e., the inherited privileges) has not been well considered in the existing access control models, whose implementation also incur large overhead. In this paper, we propose a lightweight graph-based access control model to manage the privileges between the networked processes, which provides a general solution for the pervasive applicabilities such as process inter-invoking and network-based access control.

According to the problem that the traditional building energy audit analysis and research methods are too single, which is not applicable to the analysis of specific survey objects such as campus buildings, this article has made an audit of water, electricity and gas energy consumption for campus buildings and put forward a new research program for water and electricity consumption. Compared with the traditional method, the research on electric energy consumption has improved the traditional formula and come up with a new research model for electric energy consumption in campus buildings. And the research on water energy consumption has added the analysis of daily water consumption per person, so the overall data will be accurate to the individual and make the results more accurate. Proved by the experiments, the improved model proposed in this paper can predict the energy consumption of the investigated objects more accurately.

Based on the mobile charging scheduling strategy, the Mobile Charger (MC) complements the power supply for the sensor nodes. We proposed a Mixed Charging Schedule Algorithm (MCS) based on the periodic charging scheme and the on-demand charging scheme. We add a node’s rate grouping mechanism to determine which nodes need to be recharged based on a dynamic threshold mechanism. Based on the periodic charging loops and the service station deployment mechanism, a charging loop is constructed according to guide the MC to wirelessly charge the sensor nodes. The proposed algorithm reduces the power consumption of the sensor network, improves the efficiency of MC mobile billing, and ensures the normal operation of the network.

Software Defined WiFi networks (SD-WiFi) support scalable network control functions, flexible resource allocation and changes in traffic. But the load balancing in SD-WiFi is challenging due to involvement of numerous users in the network. In this paper, we propose an efficient algorithm approach to achieve load balancing in SD-WiFi architecture. The user generated traffic arrives at WiFi access points (APs), which is classified into high prioritized (HP) flows and low prioritized (LP) flows, based on flow size and delay constraint values using support vector machine (SVM). Controllers are organized as two-tier: global controller (GC) and local controllers (LC). Markov Chain Model (MCM) is employed with two transition states as overloaded and underloaded in GC to predict future load of LCs based on the current load. The optimal underloaded LC for flow migration is selected by using Type-2 Fuzzy based Particle Swarm Optimization (TFPSO) algorithm. We conducted extensive simulation experiments to evaluate the performance of the proposed scheme using OMNeT++ simulator. The proposed scheme outperforms flow stealer scheme by a $$33\%$$ increase in throughput and $$70\%$$ in workload performance. In comparison to MPSO-CO scheme the proposed scheme exhibits better latency results.

Recently, hardware Trojan has become a major threat for integrated circuits. Most of the existing hardware Trojan detection works require golden chips or golden models for reference. However, a golden chip is extremely difficult to obtain or even does not exist. In this paper, we propose a novel hardware Trojan detection technique using unsupervised clustering techniques. The unsupervised clustering technique can obtain the structure information of the set of unlabeled ICs, and then distinguishes the suspicious ICs from the ICs under test. We formulate the unsupervised hardware Trojan detection problem into two types of detection models: partitioning-based and density-based detection model. We also propose a novel metric to determine the labels of the clusters. Compared with the state-of-the-art detection methods, the proposed technique can work in an unsupervised scenario with no need of ICs’ prior information. It does not require fabricated golden chips or golden models. We perform simulation evaluation on ISCAS89 benchmarks and FPGA evaluation on Trust-HUB benchmarks. Both evaluation results show that the proposed technique can detect infected ICs in the unsupervised scenario with a good accuracy.

In recent years, fine-grained vehicle recognition has been one of the essential tasks in Intelligent Traffic System (ITS) and has a multitude of applications, such as highway toll, parking intelligent management and vehicle safety monitoring. Fine-grained vehicle recognition is a challenging problem because of small inter-class distance and substantial sub-classes. To tackle this task, we propose a part-based model for fine-grained vehicle recognition in a weakly unsupervised manner. We also provide a part location method that locates the discriminative parts based on saliency maps which can be easily obtained by a single back-propagation pass. The advantage of the method is that the resolution of saliency maps is the same as the resolution of input images. Thus, we can locate discriminative parts efficiently and accurately. Additionally, we combine the whole-level features and part-level features and improve the accuracy of recognition up to 98.41% over 281 vehicle models in the large-scale dataset CompCars.

This paper designs and implements a new type of photovoltaic cell simulator based on LabVIEW platform, which is used to simulate the output I-V characteristics of photovoltaic cells. The main contribution of this work is the implementation of the point-by-point comparison algorithm, which is used to control the power supply of a photovoltaic cell to follow a given I-V output curve on the LabVIEW platform. The experimental results show that the photovoltaic cell simulator can achieve 96% accuracy, and can be used as a reliable tool for the simulation of the MPPT (Maximum Power Point Tracking) algorithm and the analysis of power supplies of sensor nodes in green internet of things.

Machine Learning (ML) Classification has already become one of the most commonly used techniques in many areas such as banking, medicine, spam detection and data mining applications. Often, the training of models require massive data which may contain sensitive information and the classification phase may expose the train models and the inputs from the users. Neither the models nor the train datasets and inputs should expose private information. Addressing this goal, several schemes have been proposed for privacy preserving classification. In this paper, we review those privacy preserving techiniques which applied for different machine learning classification algorithms. These algorithms conclude k-NN, SVM, Bayesian, neural networks, decision tree and etc. we sum up the comparison protocols. Finally, this work comes up with some correlative problems which are worthy to study in the future.

Nanjing has been listed as the one of the worst performers across China with respect to the high level of haze-fog, which impacts people’s health greatly. For the severe condition of haze-fog, PM2.5 is the main cause element of haze-fog pollution in China. So it’s necessary to forecast PM2.5 concentration accurately. In this paper, an artificial intelligence method is employed to forecast PM2.5 in Nanjing. At the data pre-processing stage, the main factors among the air pollutants (O3, NO2, SO2, CO, etc.) as well as meteorological parameters (pressure, wind direction, temperature, etc.) that affect PM2.5 are selected, and these factors of previous hours are as input data to predict PM2.5 concentration of next hours. Considering the air pollutants and meteorological data are typical time series data, a special recurrent neural network, which is called long short term memory (LSTM) network, is applied in this paper. To determine the amount of nodes in the hidden layer, a self-organizing method is used to automatically adjust the hidden nodes during the training phase. Finally, the PM2.5 concentrations of the next 1 h, 4 h, 8 h, and 12 h are predicted separately by using the self-organizing LSTM network based approach. The experimental result has been validated and compared to other algorithms, which reflects the proposed method performs best.

In this paper, we investigate Chen et al. biometrics-based remote user authentication scheme and find that it cannot validate the correctness of the password, complete the storage and verification of the password, and vulnerable to anonymity attacks, smart card stolen attacks and forgery attacks. To remedy these flaws, we propose an improved three-factor remote authentication scheme based on smart cards. It can implement mutual authentication and generate session keys to effectively improve security in multi-server environments. The proposed scheme can resist smart card attack, anonymity attack, forgery attack and other attacks. In addition, the proposed scheme costs $$5T_{h}$$ more compare to Chen et al. work and less computation complexity compared with other schemes.

Vehicular Ad-hoc Network (VANET) is a rapid developing application of the Internet of Things, which has totally difference with the conventional traffic network. On one hand, a VANET needs to provide end-to-end verification that has been widely concerned by researchers. On the other hand, the privacy is also very important in a VANET. Because of the openness of VANETs, conventional security measures do not effectively guarantee vehicle’s privacy. In other words, there is an urgent requirement for new security schemes to guarantee vehicular privacy in VANETs. Therefore, ensuring anonymity and authentication is a double requirement in VANETs. The both requirements seem contradictory but have to coexist. However, effective technology is few in literature and the existing anonymous authentication schemes cannot provide reverse trace mechanism when anonymity abuse happens. At the same time, these existing researches are limited to the application layer without considering the particularity of the perception layer and network layer. Without the support of the lower layer, the upper layer cannot achieve real security. Our aim is to build a lightweight anonymous authentication security system under the privacy preserving which considers the limitation of bottom layer devices and anonymity abuse problem.