Malicious Manipulations Research Articles

Convergence analysis of Iterated Best Response for a trusted computation game

We introduce a game of trusted computation in which a sensor equipped with limited computing power leverages a central computer to evaluate a specified function over a large dataset, collected over time. We assume that the central computer can be under attack and we propose a strategy where the sensor retains a limited amount of the data to counteract the effect of attack. We formulate the problem as a two player game in which the sensor (defender) chooses an optimal fusion strategy using both the non-trusted output from the central computer and locally stored trusted data. The attacker seeks to compromise the computation by influencing the fused value through malicious manipulation of the data stored on the central computer. We first characterize all Nash equilibria of this game, which turn out to be dependent on parameters known to both players. Next we adopt an Iterated Best Response (IBR) scheme in which, at each iteration, the central computer reveals its output to the sensor, who then computes its best response based on a linear combination of its private local estimate and the untrusted third-party output. We characterize necessary and sufficient conditions for convergence of the IBR along with numerical results which show that the convergence conditions are relatively tight.

Challenges in Assessing Safety Critical Autonomous Systems

Safety-critical autonomous systems like self-driving cars and fully autonomous industrial robots will integrate a number of complex technologies, each contributing to the residual failure potential of the overall system. For the validation of such systems it would be ideal to use a combination of design principles amenable to validation by formal methods and quantification of residual error probabilities. This goal, however may not always be fully achievable. We conclude that a risk assessment must be aware of those contributing factors that cannot be quantified in an analytic way, including possibilities for novel ways of malicious manipulation and unforeseen effects of interaction with the environment.

Smooth filtering identification based on convolutional neural networks

The increasing prevalence of digital technology brings great convenience to human life, while also shows us the problems and challenges. Relying on easy-to-use image editing tools, some malicious manipulations, such as image forgery, have already threatened the authenticity of information, especially the electronic evidence in the crimes. As a result, digital forensics attracts more and more attention of researchers. Since some general post-operations, like widely used smooth filtering, can affect the reliability of forensic methods in various ways, it is also significant to detect them. Furthermore, the determination of detailed filtering parameters assists to recover the tampering history of an image. To deal with this problem, we propose a new approach based on convolutional neural networks (CNNs). Through adding a transform layer, obtained distinguishable frequency-domain features are put into a conventional CNN model, to identify the template parameters of various types of spatial smooth filtering operations, such as average, Gaussian and median filtering. Experimental results on a composite database show that putting the images directly into the conventional CNN model without transformation can not work well, and our method achieves better performance than some other applicable related methods, especially in the scenarios of small size and JPEG compression.

NFD: Non-technical loss fraud detection in Smart Grid

Utility companies consistently suffer from the harassing of Non-Technical Loss (NTL) frauds globally. In the traditional power grid, electricity theft is the main form of NTL frauds. In Smart Grid, smart meters thwart electricity theft in some ways but cause more problems, e.g., intrusions, hacking, and malicious manipulation. Various detectors have been proposed to detect NTL frauds, but they either rely on user behavior analysis which requires a large amount of historical data or needs a lot of extra devices which are expensive. In this paper, a detector named NFD (NTL Fraud Detection), is proposed to detect NTL frauds with only a small amount of data and one additional device. NFD is based on Lagrange polynomial interpolation to model an adversary's behaviors, and detect a tampered meter by comparing the difference between the results. Different from existing detectors, our detector knows adversaries better than adversaries themselves. By building mathematical models of these adversaries, we can predict their behaviors which they may not be aware of by themselves. NFD makes it practical to detect NTL frauds both online and offline in Smart Grid. NFD can facilitate real-world applications before Smart Grid is fully deployed since it can serve the traditional power grid and Smart Grid simultaneously. Experimental results show the effectiveness of NFD. It can detect multiple tampered meters and multiple adversaries, as well as a single tampered meter and a single adversary. We also study how to tune the parameters used in NFD to further guide its practical usage.

Ranking scientific publications with similarity-preferential mechanism

Along with the advance of internet and fast updating of information, nowadays it is much easier to search and acquire scientific publications. To identify the high quality articles from the paper ocean, many ranking algorithms have been proposed. One of these methods is the famous PageRank algorithm which was originally designed to rank web pages in online systems. In this paper, we introduce a preferential mechanism to the PageRank algorithm when aggregating resource from different nodes to enhance the effect of similar nodes. The validation of the new method is performed on the data of American Physical Society journals. The results indicate that the similarity-preferential mechanism improves the performance of the PageRank algorithm in terms of ranking effectiveness, as well as robustness against malicious manipulations. Though our method is only applied to citation networks in this paper, it can be naturally used in many other real systems, such as designing search engines in the World Wide Web and revealing the leaderships in social networks.

The effect of malicious manipulations on prediction market accuracy

Prediction markets are a form of group decision support system which uses a market mechanism to elicit and aggregate information from large numbers of individuals. The literature recognises their potential as decision support tools, but also notes several issues of concern regarding their utility in an organisational setting. One critical concern is the possibility that prediction markets may be subject to malicious manipulation. This paper presents a field experiment which examines the effect of such manipulations on prediction market performance. We divide a sample of 72 contracts into a control group and an experimental group. Contracts in the experimental group are manipulated by a trader with a malicious motivation. The study demonstrates that manipulations do have an effect on prediction market accuracy, but that these effects are rapidly ameliorated by rational traders and shows that fear of malicious manipulation should not preclude the use of prediction markets as organisational decision making tools.

RWD based IDS for MANET

Mobile Ad-Hoc network (MANET) is a temporary infrastructure less network. This network is formed by combining some set of wireless mobile hosts [1] [5]. The host is called as a node which dynamically establishes their own network. Intrusion detection in MANETs, however, is challenging for a number of reasons. These networks change their topologies dynamically due to node mobility; lack concentration points where traffic can be analyzed for intrusions; utilize self-configuring multi-party infrastructure protocols that are susceptible to malicious manipulation; and rely on wireless communications channels that provide limited bandwidth .To overcome these constraints, researchers have proposed a number of intrusion detection approaches specifically for MANETs. Intrusion detection is a process of identifying and responding to malicious activity targeted at computing and networking resources. In this IDS architecture multilayer specification based detection engine is used along with Random Walker Detector (RWD) [1] [16]. It randomly traverses a network and find outs that on which node which attack is occurred. If there is attacks it performs re-routing.

Call for Papers: Issue 1/2017

Human information behavior (HIB) is a multi-disciplinary field at the intersection of different scientific disciplines, including information systems (IS), psychology, information science, sociology, and neuroscience. It is concerned with explaining and predicting human behaviors towards information. This includes behaviors relating to acquiring, interpreting, using, and sharing information, in addition to behaviors relating to limiting, filtering, and stopping the information intake. Information pathologies are of special relevance to HIB research. They describe misguided and malicious behaviors that cause relevant information not to be produced, absorbed, used, or shared. Examples are information avoidance, information hiding, and the malicious manipulation of information. HIB topics bring together researchers with diverse backgrounds, epistemological stances, and research methods. The unit of observation is the individual, the group, the organization, and the society as a whole. This special issue seeks contributions that explicitly address the cognitive and affective processes that relate to information behaviors and the types of information behaviors evident in the IS literature. Furthermore, it seeks contributions dealing with the information construct itself, in addition to issues relating to information quality, information relevance, and information value. This BISE journal special issue is open to contributions regardless of the scientific methodology chosen. Innovative approaches such as the use of neurophysiology (NeuroIS) are highly appreciated; but the special issue is not restricted to these methods. The special issue welcomes IS research contributions which explicitly deal with HIB aspects related to the development and use of IS. Contributions from research and business practice on the following (and related) topics are invited:

Image Authentication using RSA and Chaotic Map

Authentication is basically a way by which sender and recipient must prove their identities to each other. Image authentication verifies the originality of an image by detecting malicious manipulations. Reliable image authentication technology must be able to protect an image from time it was first produced until the final stage of use. In this paper we describe an effective technique for image authentication which can prevent malicious manipulations by using RSA and Baker’s map. Chaos refers to a type of complex dynamical behavior that possesses some special features such as being extremely sensitive to small variations in initial conditions. Moreover this is certificateless-based scheme and allows people to verify the signature without the certificate. For this reason, we do not need the certificate authority to store and manage user’s certificates. In 2014, Chin-Chen Chang, Chin-Yu Sun, and Shih-Chang Chang has given a similar designed certificateless-based signature scheme based on RSA operations; however, their scheme is modified and improved by using it on image rather than text [2]. General Terms Chaotic map, RSA, Digital Signature, Random number, MD5. SHA-1.

RST Invariant Fragile Watermarking for 2D Vector Map Authentication

D vector maps are very important data in geographical information systems (GISs). It is considered to be content for which verification of integrity and authentication are urgently required. Using a marking technique and a rotation, scaling and translation (RST) invariant watermark generation method, we propose a scheme that detects and locates malicious attacks with high accuracy while providing the RST invariance property. In particular, we propose setting a unique mark on each feature, calculating the authentication watermark for each spatial feature group by folding the hash results of the differences of the log-radiuses, and embedding the watermark using a RST invariant watermarking method. While the watermark ensures both the sensitivity to malicious manipulations and the RST invariance property, the mark of each feature provides superior accuracy of tamper localization. Moreover, this algorithm is immune to feature rearrangement and vertex reversing operations. Theoretical analysis and experimental results are provided to demonstrate the effectiveness of our method.

Image Authentication by Content Preserving Robust Image Hashing Using Local and Global Features

Image hashing technique constructs a short sequence from the image to represent its contents. This method proposes an image hash which is generated from Haralick and MOD-LBP features along with luminance and chrominance, which are computed from Zernike moments. Sender generates the hash from image features and attaches it with the image to be sent. The hash is analyzed at the receiver to examine the authenticity of the image. The method detects image forgery and locates the forged regions of the image. The proposed hash is robust to common content preserving modifications and sensitive to malicious manipulations.

Ranking scientific publications: the effect of nonlinearity.

Ranking the significance of scientific publications is a long-standing challenge. The network-based analysis is a natural and common approach for evaluating the scientific credit of papers. Although the number of citations has been widely used as a metric to rank papers, recently some iterative processes such as the well-known PageRank algorithm have been applied to the citation networks to address this problem. In this paper, we introduce nonlinearity to the PageRank algorithm when aggregating resources from different nodes to further enhance the effect of important papers. The validation of our method is performed on the data of American Physical Society (APS) journals. The results indicate that the nonlinearity improves the performance of the PageRank algorithm in terms of ranking effectiveness, as well as robustness against malicious manipulations. Although the nonlinearity analysis is based on the PageRank algorithm, it can be easily extended to other iterative ranking algorithms and similar improvements are expected.

A Hybrid Scheme for Authenticating Scalable Video Codestreams

A scalable video coding (SVC) codestream consists of one base layer and possibly several enhancement layers. The base layer, which contains the lowest quality and resolution images, is the foundation of the SVC codestream and must be delivered to recipients, whereas enhancement layers contain richer contour/texture of images in order to supplement the base layer in resolution, quality, and temporal scalabilities. This paper presents a novel hybrid authentication (HAU) scheme. The HAU employs both cryptographic authentication and content-based authentication techniques to ensure integrity and authenticity of the SVC codestreams. Our analysis and experimental results indicate that the HAU is able to detect malicious manipulations and locate the tampered image regions while is robust to content-preserving manipulations for enhancement layers. Although our focus in this paper is on authenticating H.264/SVC codestreams, the proposed technique is also applicable to authenticate other scalable multimedia contents such as MPEG-4 fine grain scalability and JPEG2000 codestreams.

Adaptive Semi-Fragile Watermarking Based on Complete Quantization and Image Contents

Complete quantization principle is proposed in this paper, based on which, an adaptive semi-fragile watermarking algorithm based on image contents is proposed. The algorithm uses the low frequency sub-band to construct watermarking related to the content of image, and then uses just noticeable difference of wavelet domain to embed watermarking by means of complete quantitative principle. Experiments show this algorithm can classify malicious and non-malicious manipulations accurately.

Image Authentication System using LDPC and Watermarking Technique

Now a days, the rate of using digital image is increasing exponentially because of its low cost and easy manipulation property. It is obvious that digital images captured from CCD cameras or other digital image devices have to be analyzed and determined before processing it, to keep the integrity and ensure the accuracy and reliability of the digital image. We need some sort of robust methods and standard procedures that guarantee and strengthen the authenticity of digital image. Image authentication is a technique that analyzes a digital image and determines whether it is altered or not. Image authentication technique is very useful for various organizations such as health care, law enforcement agencies and insurance sector etc. Image authentication is also important in content delivery via untrusted intermediaries, such as peer-to-peer (P2P) file sharing. Untrusted intermediaries might tamper the contents of image. Distinguishing the legitimate diversity of encodings from malicious manipulation is the challenge addressed in this paper. We developed a LDPC and watermark based new approach for image authentication. With the help of this approach we can authenticate images effectively. In our technique, we provide LDPC quantized image projection and the Encrypted image as authentication data. As well as watermark image that was embedded into original image to identify legitimate or illegitimate state of image authentication system. These data can be correctly decoded only with the help of an authentic image as side information. This technique provides the desired robustness against legitimate encoding alteration, while detecting illegitimate variations.

Implementation of Technique for Image Authentication using Regular LDPC Codes

In this paper, we propose an approach using encryption technique and LDPC source coding for the image authentication problem. Image authentication is important in content delivery via untrusted intermediaries, such as peer-topeer (P2P) file sharing. Many differently encoded versions of the original image might exist. In addition, intermediaries might tamper with the contents. Distinguishing legitimate diversity from malicious manipulations is the challenge addressed in this research. The key idea is to provide a Slepian-Wolf encoded quantized image projection as authentication data which is again encrypted using a secret key cryptography before ready to send. This can be correctly decoded with the help of an authentic image as side information. This mechanism provides the desired robustness against legitimate variations while detecting illegitimate modification. The decoder incorporating expectation maximization (EM) algorithms can authenticate images which have undergone contrast, brightness and even warping adjustments. Our novel authentication system also others tampering localization by using inference over a factor graph that represents tampering models.

Watermarking for multi-stage encoded image authentication

Multi-stage encoded images generally provide multiple resolutions that have advantages in image transmission and storage. Several image encoders such as JPEG, SPIHT, EZW, JPEG2000, etc. support multi-stage encoding. However, any one of multi-stage encoded images should also be authentic to ensure transmission security. In this paper, a watermarking-based authentication scheme for multi-stage encoded images is proposed. The image authentication message is extracted from the image content and then embedded into images according to the multi-stage encoding. The experimental results show the validity of the proposed multi-stage image authentication scheme. A malicious operation can be detected within a multi-stage encoded image in which any significant tampering can be detected in the earlier authentication stages and any slight tampering can be detected in the later authentication stages. Furthermore, the proposed scheme can distinguish acceptable operations from malicious manipulations and the positions of the tampered areas can be located.

Image Retrieval using Perceptual Hashing

This paper proposes a new perceptual hashing based approach for Image Retrieval. Information in form of images is increasing at a very fast rate therefore there is need for faster and effective retrieval of specific image from image repositories. Content Based Image Retrieval (CBIR) is considered most efficient to search large image collections. The proposed method uses perceptual hash function which calculates similar hash value for similar images. Finally, using an adequate distance or similarity function to compare two perceptual hash values, it can be decided whether two images are perceptually different or not. Perceptual image hash functions can be used e.g. for the identification, authentication or integrity verification of images. In this paper we also propose a hash algorithm that is robust and secure to non malicious manipulation and sensitive to the malicious tampering. Keywords - Content based Image Retrieval (CBIR), Perceptual Hash algorithm, Image Retrieval using Perceptual Hashing, Query by Image Content (QBIC).

Secure healthcare data sharing among federated health information systems

The current trend in designing health information systems is to apply federated architectures to integrate existing systems. This exacerbates the security guarantees that such systems are required to satisfy and demands the introduction of advanced methods to deal with security. This paper aims at describing how federated health information systems can offer security properties by adopting proper mechanisms to protect the exchanged data and the provided functionalities from malicious manipulations. We have experimentally evaluated the costs in terms of performance penalty related to the introduction of security mechanisms within the proposed solution.

Robust Image Alignment for Tampering Detection

The widespread use of classic and newest technologies available on Internet (e.g., emails, social networks, digital repositories) has induced a growing interest on systems able to protect the visual content against malicious manipulations that could be performed during their transmission. One of the main problems addressed in this context is the authentication of the image received in a communication. This task is usually performed by localizing the regions of the image which have been tampered. To this aim the aligned image should be first registered with the one at the sender by exploiting the information provided by a specific component of the forensic hash associated to the image. In this paper we propose a robust alignment method which makes use of an image hash component based on the Bag of Features paradigm. The proposed signature is attached to the image before transmission and then analyzed at destination to recover the geometric transformations which have been applied to the received image. The estimator is based on a voting procedure in the parameter space of the model used to recover the geometric transformation occurred into the manipulated image. The proposed image hash encodes the spatial distribution of the image features to deal with highly textured and contrasted tampering patterns. A block-wise tampering detection which exploits an histograms of oriented gradients representation is also proposed. A non-uniform quantization of the histogram of oriented gradient space is used to build the signature of each image block for tampering purposes. Experiments show that the proposed approach obtains good margin of performances with respect to state-of-the art methods.