The ISC International Journal of Information Security

Abstract The Tor network is probably one of the most popular online anonymity systems in the world. It has been built based on the volunteer relays from all around the world. It has a strong scientific basis which is structured very well to work in low latency mode that makes it suitable for tasks such as web browsing. Despite the advantages, the low latency also makes Tor insecure against timing and traffic analysis attacks, which are the most dominant attacks on Tor network in recent past years. In this paper, first all kinds of attacks on Tor network will be classified and then timing and traffic analysis attacks will be described in more details. Then we present a new circuit scheduling for Tor network in order to preserve two properties, fairness and randomness. Both properties are trying to make pattern and timing analysis attacks more difficult and even in some cases impractical. Our scheduler distorts timing patterns and size of packets in a random way (randomness) without imposing artificial delays or paddings (fairness). Finally, by using our new scheduler, one of the most powerful attacks in this area is debilitated, and by it is shown that analyzing traffic patterns and size of packets will be more difficult to manage.

Abstract A certificateless (CL) signcryption scheme is a cryptographic primitive that provides user authentication and message confidentiality at the same time. CL signcryption schemes (as a type of certificateless encryption scheme) have solved problems concerning malicious server presentation, and the server who issues users' partial private keys and certificates cannot obtain users' signing keys. Therefore, the CL signcryption scheme is an excellent choice for protecting users' signing keys and providing user authentication and message confidentiality. Moreover, signcryption schemes have lower computational costs than signature and encryption schemes.
The present study presents a short and efficient CL signcryption scheme based on the hyperelliptic curve (HC). Applying HC as the calculation base for designing the presented CL signcryption scheme reduces key-length from 160 bits to 80. The presented CL signcryption scheme is shorter than other recently-proposed ones with regard to communication overhead with its less than one-third shorter length compared to the shortest of the others. Moreover, it is more efficient than other recently-proposed CL signcryption schemes in the user-side computational cost, including the \textit{key generation} and \textit{user key generation} phases that have been halved in total. Finally, the security of the presented CL signcryption scheme was analyzed in the random oracle (RO) model based on the hardness of the point factorization problem (PFP) on HC.

Abstract Today, the Internet of Things (IoT) is one of the emerging technologies that enable the connection and transfer of information through communication networks. The main idea of the IoT is the widespread presence of objects such as mobile devices, sensors, and RFID. With the increase in traffic volume in urban areas, the existing intelligent urban traffic management system based on IoT can be vital. Therefore, this paper focused on security in urban traffic based on using RFID. In our scheme, RFID tags chose as the purpose of this article. We, in this paper, present a mutual authentication protocol that leads to privacy based on hybrid cryptography. Also, an authentication process with RFID tags is proposed that can be read at high speed. The protocol has attempted to reduce the complexity of computing. At the same time, the proposed method can withstand attacks such as spoofing of tag and reader, tag tracking, and replay attack.

Abstract The present paper is intended to present a robust multiplicative video watermarking scheme. In this regard, the video signal is segmented into 3-D blocks like cubes, and then, the 3-D wavelet transform is applied to each block. The low frequency components of the wavelet coefficients are then used for data embedding to make the process robust against both malicious and unintentional attacks. The hidden message is inserted through multiplying/dividing these coefficients by a constant parameter which controls the power of the watermark. The watermark extraction relies on a maximum likelihood-based procedure, observing the distribution of the watermarked coefficients. The performance of the proposed scheme has been verified via simulations and found to be superior to some of the well-known existing video watermarking methods.

Abstract Confidentiality, unforgeability, and public verifiability are essential for secure multi-party communications. These communications play a vital role in real-world applications such as decentralized financial transactions, e-commerce, cloud computing, and web services, where authentication and privacy preservation are very important. In conventional cryptosystems, individual signcryption performed by each participant significantly increases the unsigncryption cost for the receiver. Multi-signcryption offers an efficient alternative by allowing multiple signers to jointly signcrypt a single message. This paper proposes a novel certificateless multi-signcryption scheme that eliminates the certificate management problem of traditional public key infrastructures and avoids the key escrow problem of identity-based cryptography. To reduce the computational cost associated with bilinear pairings over elliptic curves, the proposed scheme is designed in a pairing-free environment. This scheme achieves constant-time verification in the unsigncryption phase and is independent of the number of signers. Security is formally proven under the hardness assumptions of the Elliptic Curve Computational Diffie–Hellman Problem (ECCDHP) and the Elliptic Curve Discrete Logarithm Problem (ECDLP). The proposed scheme ensures confidentiality, unforgeability, and public verifiability, and it attains significantly lower computational costs than existing schemes. Hence, the proposed scheme can be used for secure group communications in resource-constrained environments where high performance is essential. 

Abstract In terms of security, MDS matrices are one of the best choices for diffusion layer of block ciphers. However, as these matrices grow in size, their software implementation becomes a challenge. In this paper, to benefit from the properties of MDS matrices and avoid the mentioned challenge, we use 4*4 MDS matrices to build some 16*16 matrices with low number of zero elements. We show that if these matrices are used as diffusion layer of software-based SPN structures, the resulting block ciphers have similar properties as AES in software implementation complexity (i.e. the number of required CPU instructions) and resistance against linear and differential attacks. Moreover, the best impossible differential and square distinguishers for the proposed 16*16 structures have similar length as SPN structures with 16*16 MDS matrices. Thus, the new structures outperform AES with respect to the impossible differential and square attacks. Additionally, we show that if the proposed SPN structure uses the AES key schedule, its results for the differential related-key attacks are better than those for AES. We also extend the idea and use 4*4 MDS matrices to design 24*24 and 32*32 matrices with acceptable properties for SPN structure design. Finally, we extend the idea to propose some matrices for Feistel structures with SP-type F-functions. We show that the resulting structures are more secure than the improved type-II GFS.

Abstract One of the security issues in data outsourcing is the enforcement of the data owner’s access control policies. This includes some challenges. The first challenge is preserving confidentiality of data and policies. One of the existing solutions is encrypting data before outsourcing which brings new challenges; namely, the number of keys required to access authorized resources, efficient policy updating, write access control enforcement, overhead of accessing/processing data at the user/owner side. Most of the existing solutions address only some of the challenges, while imposing high overhead on both owner and users. Though, policy management in the Role-Based Access Control (RBAC) model is easier and more efficient due to the existence of role hierarchical structure and role inheritance; most of the existing solutions address only enforcement of policies in the form of access control matrix. In this paper, we propose an approach to enforce RBAC policies on encrypted data outsourced to a service provider. We utilize Chinese Remainder Theorem for key management and role/permission assignment. Efficient user revocation, efficient role hierarchical structure updating, availability of authorized resources for users of new roles, and enforcement of write access control policies as well as static separation of duties, are of advantages of the proposed solution.

Abstract Electronic health record (EHR) system facilitates integrating patients' medical information and improves service productivity. However, user access to patient data in a privacy-preserving manner is still challenging problem. Many studies concerned with security and privacy in EHR systems. Rezaeibagha and Mu [1] have proposed a hybrid architecture for privacy-preserving accessing patient records in a cloud system. In their scheme, encrypted EHRs are stored in multiple clouds to provide scalability and privacy. In addition, they considered a role-based access control (RBAC) such that for any user, an EHR access policy must be determined. They also encrypt the EHRs by the public keys of all users. So, for a large amount of EHRs, this scheme is not efficient. Furthermore, using RBAC for access policy makes the policy changing difficult. In their scheme, users cannot search on encrypted EHRs based on diseases and some physicians must participate in the data retrieval by a requester physician. In this paper, we address these problems by considering a ciphertext-policy attribute-based encryption (CP-ABE) which is conceptually closer to the traditional access control methods such as RBAC. Our secure scheme can retrieve encrypted EHR based on a specific disease. Furthermore, the proposed scheme guarantees the user access control and the anonymity of the user or data owner during data retrieval. Moreover, our scheme is resistant against collusion between unauthorized retrievers to access the data. The analysis shows that our scheme is secure and efficient for cloud-based EHRs.

Abstract Nowadays Radio Frequency Identification (RFID) systems have appeared in lots of identification and authentication applications. In some sensitive applications, providing secure and confidential communication is very important for end-users. To this aim, different RFID authentication protocols have been proposed, which have tried to provide security and privacy of RFID users. In this paper, we analyze the privacy of two recently proposed RFID authentication protocols in 2012 and 2013. We present several traceability attacks including traceability, backward traceability and forward traceability against the first protocol. We also show that, the second protocol not only suffers from Denial-of-Service (DoS) attack, but also it is vulnerable to traceability and backward traceability attacks. We present our privacy analysis based on a well-known formal RFID privacy model which has been proposed by Ouafi and Phan in 2008. Then, in order to overcome the weaknesses, we apply some modifications on these protocols and propose two modified versions.

Abstract The products of graphic design applications leave behind traces of digital information which can be used during a digital forensic investigation in cases where counterfeit documents have been created. This paper analyzes the digital forensics involved in the creation of counterfeit documents. This is achieved by first recognizing the digital forensic artifacts left behind from the use of graphic design applications, and then analyzing the files associated with these applications. When analyzing digital forensic artifacts generated by an application, the specific focus is on determining whether the graphic design application was installed, whether the application was used, and determining whether an association can be made between the application’s actions and such a digital crime. This is accomplished by locating such information from the registry, log files and prefetch files. The file analysis involves analyzing files associated with these applications for file signatures and metadata. In the end it becomes possible to determine if a system has been used for creating counterfeit documents or not.

Abstract Quantitative and model-based prediction of security in the architecture design stage facilitates early detection of design faults hence reducing modification costs in subsequent stages of software life cycle. However, an important question arises with respect to the accuracy of input parameters. In practice, security parameters can rarely be estimated accurately due to the lack of sufficient knowledge. This inaccuracy is ignored in most of the existing evaluation methods. The aim of this paper is to explicitly consider parameter uncertainty in the software security evaluation process. In particular, we use the Dempster-Shafer theory of evidence to formulate the uncertainties in input parameters and determine their effects on output measures. In the proposed method, security attacks are expressed using UML diagrams (i.e., misuse case and mal-activity diagrams) and security parameters are specified using the SecAM profile. UML/SecAM models are then transformed into attack trees, which allow quantifying the probability of security breaches. The applicability of the method is validated by a case study on an online marketing system.

Abstract The aim of phishing is tracing the users' s private information without their permission by designing a new website which mimics the trusted website. The specialists of information technology do not agree on a unique definition for the discriminative features that characterizes the phishing websites. Therefore, the number of reliable training samples in phishing detection problems is limited. Moreover, among the available training samples, there are abnormal samples that cause classification error. For instance, it is possible that there are phishing samples with similar features to legitimate ones and vice versa. A supervised feature extraction method, called weighted feature line embedding, is proposed in this paper to solve these problems. The proposed method virtually generates training samples by utilizing the feature line metric. Hence, it can solve the small sample size problem. Moreover, by assigning appropriate weights to each pair of feature points, it corrects the undesirable quality of abnormal samples. The features extracted by our method improve the performance of phishing website detection specially by using small training sets.

Abstract Attribute-based encryption (ABE) is one of the recommended tools to secure real systems like the Internet of Things (IoT). Almost all the ABE schemes utilize bilinear map operations, known as pairings. The challenge with these schemes is that performing pairings results in high computation costs and IoT devices are typically resource-constrained, so, efficient pairing-free ABE schemes have been proposed to solve this issue. These schemes utilize classical cryptographic operations instead of heavy bilinear pairings. Recently, two pairing-free ciphertext-policy attribute-based encryption schemes have been proposed (by Das et al. and Sowjanya et al.). According to their claims, their schemes are secure against collusion attacks and provide indistinguishability in a selective-set security model. The first scheme also has been claimed to be secure against forgery attacks. In this paper, we show that the first scheme is vulnerable to ciphertext-only, collusion between four or more data users with specific features, and forgery attacks. We also show that the second scheme is vulnerable to a key recovery attack, which can lead to a collusion attack. So, even though they are highly efficient, they have some security vulnerabilities that can violate the claims of the authors.

Abstract Abstract- With the advancement and development of computer network technologies, the way for intruders has become smoother; therefore, to detect threats and attacks, the importance of intrusion detection systems (IDS) as one of the key elements of security is increasing. One of the challenges of intrusion detection systems is managing of the large amount of network traffic features. Removing unnecessary features is a solution to this problem. Using machine learning methods is one of the best ways to design an intrusion detection system. Focusing on this issue, in this paper, we propose a hybrid intrusion detection system using the decision tree and support vector machine (SVM) approaches. In our method, the feature selection is initially done by the C5.0 decision tree pruning, and then the features with the least predictor importance value are removed. After removing each feature, the least square support vector machine (LS-SVM) is applied. The set of features having the highest surface area under the Receiver Operating Characteristic (ROC) curve for LS-SVM are considered as final features. The experimental results on two KDD Cup 99 and UNSW-NB15 data sets show that the proposed approach improves true positive and false positive criteria and accuracy compared to the best prior work.

Abstract The aim of image steganalysis is to detect the presence of hidden messages in stego images. We propose a blind image steganalysis method in Contourlet domain and then show that the embedding process changes statistics of Contourlet coefficients. The suspicious image is transformed into Contourlet space, and then the statistics of Contourlet subbands coefficients are extracted as features. We use absolute Zernike moments and characteristic function moments of Contourlet subbands coefficients of the image to distinguish between the stego and non-stego images. Absolute Zernike moments are used to examine the randomness in the test image and characteristic function moments of Contourlet coefficients is used to form our feature set that can catch the changes made to the histogram of Contourlet coefficients. These features are fed to a nonlinear SVM classifier with an RBF kernel to distinguish between cover and stego images. We show that the embedding process distorts statistics of Contourlet coefficients, leading to detection of stego images. Experimental results confirm that the proposed features are highly sensitive to the change made by the embedding process. These results also reveal advantage of the proposed method over its counterpart steganalyzers, in cases of five popular JPEG steganography techniques.

Abstract With the spread of information technology in human life, data protection is a critical task. On the other hand, malicious programs are developed, which can manipulate sensitive and critical data and restrict access to this data. Ransomware is an example of such a malicious program that encrypts data, restricts users' access to the system or their data, and then request a ransom payment. Many types of research have been proposed for ransomware detection. Most of these methods attempt to identify ransomware by relying on program behavior during execution. The main weakness of these methods is that it is not explicit how long the program should be monitored to show its real behavior. Therefore, sometimes, these researches cannot detect ransomware early. In this paper, a new method for ransomware detection is proposed that does not need executing the program and uses the PE header of the executable file. To extract effective features from the PE header file, an image is constructed based on PE header. Then, according to the advantages of Convolutional Neural Networks in extracting features from images and classifying them, CNN is used. The proposed method achieves high detection rates. Our results indicate the usefulness and practicality of our method for ransomware detection.

Abstract A convertible limited (multi-) verifier signature (CL(M)VS) provides controlled verifiability and preserves the privacy of the signer. Furthermore, limited verifier(s) can designate the signature to a third party or convert it into a publicly verifiable signature upon necessity. In this proposal, we first present a generic construction of convertible limited verifier signature (CLVS) into which the existing secure CLVS schemes fit. Afterwards, we extend this generic construction to address the unsolved question of designing an efficient construction with more than two limited verifiers. To this effect, two generic CLMVS constructions are presented, which are proven to be efficient in that they generate a unique signature for more than two limited verifiers. Given the first generic construction, each limited verifier checks the validity of the signature solely, while in the second, cooperation of all limited verifiers is imperative. Thereupon, on the ground of our second generic construction, we present the first pairing-based CLMVS scheme secure in the standard model, which is of a strong confirmation property as well. Finally, we employ the proposed CLMVS scheme for one limited verifier (CLVS) so as to design a new electronic voting protocol.

Abstract Distributed Denial of Service (DDoS) attacks are of two kinds viz. high-rate DDoS (HRDDoS) attacks and low-rate DDoS (LRDDoS) attacks. A shrew attack is a LRDDoS attack that can prove to be more harmful than a HRDDoS attack since they are not easily noticeable and are stealthy. They cause TCP flows to attain near-zero throughput by sending attack pulses of very short bursts synchronized with the TCP retransmission timeout (RTO) value. Consequently, it compels the TCP packets to be dropped whenever it tries to retransmit again after the timeout. Thus, it may endanger the victim systems if not detected for a long time and reduce the overall quality of services without being noticed. In this paper, we perform the analysis of the network traffic based on a statistical approach where the deviation in the behavior of the flows is analyzed based on the packets sent during the normal and attack conditions. To do this, we determine the participation of a flow in congestion and its adherence to the legitimate TCP-compliant nature during attack conditions based on a priority determiner D. The shrew attack flows exhibit higher values of $D$ as they do not adhere to the TCP compliance and tend to contribute to more congestion to disrupt a server. This nature of attack flows enables us to filter them based on the values of $D$ and mitigate them by blocking these flows. The experimental results on various scenarios demonstrated high accuracy to substantiate the efficacy of the proposed method.

Abstract Using generative models to produce unlimited synthetic samples is a popular replacement for database sharing. Generative Adversarial Network (GAN) is a popular class of generative models which generates synthetic data samples very similar to real training datasets. However, GAN models do not necessarily guarantee training privacy as these models may memorize details of training data samples. When these models are built using sensitive data, the developers should ensure that the training dataset is appropriately protected against privacy leakage. Hence, quantifying the privacy risk of these models is essential. To this end, this paper focuses on evaluating the privacy risk of publishing the generator network of GAN models. Specially, we conduct a novel generator white-box membership inference attack against GAN models that exploits accessible information about the victim model, i.e., the generator’s weights and synthetic samples, to conduct the attack. In the proposed attack, an auto-encoder is trained to determine member and non-member training records. This attack is applied to various kinds of GANs. We evaluate our attack accuracy with respect to various model types and training configurations. The results demonstrate the superior performance of the proposed attack on non-private GANs compared to previous attacks in white-box generator access. The accuracy of the proposed attack is 19% higher on average than similar work. The proposed attack, like previous attacks, has better performance for victim models that are trained with small training sets.

Abstract Backdoor attacks pose a significant threat to deep learning systems by injecting hidden malicious behavior to the model while preserving high accuracy on clean data. Such attacks are particularly dangerous in scenarios where users rely on pre-trained models or outsource training to untrusted parties. In this work, we propose a practical defense strategy that assumes no knowledge of the backdoor trigger or the training process, relying on a small trusted clean dataset. Our method introduces a two-stage pipeline: First, we aggregate predictions from multiple potentially compromised models to train an intermediate Teacher-Aggregation (TA) model; then, we distill this knowledge into a compact light-weight student model. This multi-stage approach effectively alleviates backdoor effects while preserving clean accuracy. Experimental results on MNIST and CIFAR-10 demonstrate that our method significantly reduces the Attack Success Rate (ASR)—to approximately 0.1% on MNIST and 2.6% on CIFAR-10—outperforming baseline ensemble defenses. Furthermore, our lightweight student model is suitable for edge deployment, providing a generic and scalable defense that remains robust under minimal assumptions, making it well-suited for real-world applications in adversarial environments. Our code is available at: https://github.com/mr-pylin/backdoor-toolbox

Abstract In smart grids, messages exchanged between service providers and smart meters should be authenticated and confidential to prevent threats due to their insecurity. Hence, it is imperative to design a secure authentication and key exchange scheme to create a session key for secure and authenticated transmission of messages. In this paper, we show that the mutual authentication and key establishment protocol presented by Sureshkumar et al. in 2020, which is based on elliptic curve cryptography (ECC), fails to satisfy forward secrecy, while they claimed that it provides perfect forward secrecy. In addition, it will be demonstrated that it is not secure against stolen database attacks of a service provider, which leads to the smart meter impersonation attack and session key exposure.
Moreover, we prove that it fails to achieve security against known sessionspecific temporary information attacks. Next, an improved authenticated key establishment protocol to address these vulnerabilities has been proposed. Then, we analyze its security with informal and formal methods, such as BurrowAbadi-Needham (BAN) logic and ProVerif. Finally, with the comparison of security features and computation and communication overhead, we show that it outperforms baseline papers.

Abstract Standard face recognition algorithms that use standard feature extraction techniques always suffer from image performance degradation. Recently, singular value decomposition and low-rank matrix are applied in many applications,
including pattern recognition and feature extraction. The main objective of this research is to design an efficient face recognition approach by combining many techniques to generate efficient recognition results. The implemented face
recognition approach is concentrated on obtaining significant rank matrix via applying a singular value decomposition technique. Measures of dispersion are used to indicate the distribution of data. According to the applied ranks, there
is an adequate reasonable rank that is important to reach via the implemented procedure. Interquartile range, mean absolute deviation, range, variance, and standard deviation are applied to select the appropriate rank. Rank 24, 12, and 6
reached an excellent 100% recognition rate with data reduction up to 2 : 1, 4 : 1 and 8 : 1 respectively. In addition, properly selecting the adequate rank matrix is achieved based on the dispersion measures. Obtained results on standard face databases verify the efficiency and effectiveness of the implemented approach.

Abstract As protection of web applications are getting more and more important every day, CAPTCHAs are facing booming attention both by users and designers. Nowadays, it is well accepted that using visual concepts enhance security and usability of CAPTCHAs. There exist few major different ideas for designing image CAPTCHAs. Some methods apply a set of modifications such as rotations to the original image saved in the data base, to make the CAPTCHA more secure.
In this paper, two different approaches for designing image based CAPTCHAs are introduced. The first one _ which is called Tagging image CAPTCHA _ is based on pre-tagged images, using geometric transformations to increase security, and the second approach tries to enhance the first one by eliminating the use of tags and relying on semantic visual concepts. In fact, recognition of upright orientation is used as a visual cue. The usability of the proposed approaches is verified using human subjects. An estimation of security is also obtained by different kinds of attacks. Further studies are done on the proposed transformations and also on the properness of each original image for each approach. Results suggest a practical Semantic Image CAPTCHA which is usable and secure compared to its peers.

Abstract We propose a new online sortition protocol which is decentralized. We argue that our protocol has safety, fairness, randomness, non-reputation and openness properties. Sortition is a process that makes random decision and it is used in competitions and lotteries to determine who is the winner. In the real world, sortition is simply done using a lottery machine and all the participant can be sure about the safety, fairness, randomness, non-reputation, and openness properties. But how we can do the sortition in virtual world such that it satisfies the desired properties? The idea is decentralization. Using cryptography notions, we provide a protocol where all agents participate in computing the winner of sortition. Our proposed protocol is novel and completely differs from other sortition protocols and also it is decentralized. It is simple and easily can be implemented and find the commercial use for those markets who want to give present to their customers in a fair and clear manner.

Abstract Impossible differential cryptanalysis, the extension of differential cryptanalysis, is one of the most efficient attacks against block ciphers. This cryptanalysis method has been applied to most of the block ciphers and has shown significant results. Using structures, key schedule considerations, early abort, and pre-computation are some common methods to reduce complexities of this attack. In this paper, we present a new method for decreasing the time complexity of impossible differential cryptanalysis through breaking down the target key space into subspaces, and extending the results on subspaces to the main target key space. The main advantage of this method is that there is no need to consider the effects of changes in the values of independent key bits on each other. Using the 14-round impossible differential characteristic observed by Boura et al. at ASIACRYPT 2014, we implement this method on 23-round LBlock and demonstrate that it can reduce the time complexity of the previous attacks to 2^71.8 23-round encryptions using 2⁵⁹ chosen plaintexts and 2 ⁷³ blocks of memory.