Source author record

Yogachandran Rahulamathavan

Yogachandran Rahulamathavan appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Cryptography and Security Artificial Intelligence Computer Vision Machine Learning Networking and Internet Architecture

Catalog footprint

What is connected

5works

5topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2022arXiv

Deep Learning for Encrypted Traffic Classification and Unknown Data Detection

Despite the widespread use of encryption techniques to provide confidentiality over Internet communications, mobile device users are still susceptible to privacy and security risks. In this paper, a new Deep Neural Network (DNN) based user activity detection framework is proposed to identify fine grained user activities performed on mobile applications (known as in-app activities) from a sniffed encrypted Internet traffic stream. One of the challenges is that there are countless applications, and it is practically impossible to collect and train a DNN model using all possible data from them. Therefore, in this work we exploit the probability distribution of DNN output layer to filter the data from applications that are not considered during the model training (i.e., unknown data). The proposed framework uses a time window based approach to divide the traffic flow of an activity into segments, so that in-app activities can be identified just by observing only a fraction of the activity related traffic. Our tests have shown that the DNN based framework has demonstrated an accuracy of 90% or above in identifying previously trained in-app activities and an average accuracy of 79% in identifying previously untrained in-app activity traffic as unknown data when this framework is employed.

preprint2022arXiv

Privacy-preserving Similarity Calculation of Speaker Features Using Fully Homomorphic Encryption

Recent advances in machine learning techniques are enabling Automated Speech Recognition (ASR) more accurate and practical. The evidence of this can be seen in the rising number of smart devices with voice processing capabilities. More and more devices around us are in-built with ASR technology. This poses serious privacy threats as speech contains unique biometric characteristics and personal data. However, the privacy concern can be mitigated if the voice features are processed in the encrypted domain. Within this context, this paper proposes an algorithm to redesign the back-end of the speaker verification system using fully homomorphic encryption techniques. The solution exploits the Cheon-Kim-Kim-Song (CKKS) fully homomorphic encryption scheme to obtain a real-time and non-interactive solution. The proposed solution contains a novel approach based on Newton Raphson method to overcome the limitation of CKKS scheme (i.e., calculating an inverse square-root of an encrypted number). This provides an efficient solution with less multiplicative depths for a negligible loss in accuracy. The proposed algorithm is validated using a well-known speech dataset. The proposed algorithm performs encrypted-domain verification in real-time (with less than 1.3 seconds delay) for a 2.8\% equal-error-rate loss compared to plain-domain verification.

preprint2020arXiv

Scalar Product Lattice Computation for Efficient Privacy-preserving Systems

Privacy-preserving applications allow users to perform on-line daily actions without leaking sensitive information. Privacy-preserving scalar product is one of the critical algorithms in many private applications. The state-of-the-art privacy-preserving scalar product schemes use either computationally intensive homomorphic (public-key) encryption techniques such as Paillier encryption to achieve strong security (i.e., 128-bit) or random masking technique to achieve high efficiency for low security. In this paper, lattice structures have been exploited to develop an efficient privacy-preserving system. The proposed scheme is not only efficient in computation as compared to the state-of-the-art but also provides high degree of security against quantum attacks. Rigorous security and privacy analyses of the proposed scheme have been provided along with a concrete set of parameters to achieve 128-bit and 256-bit security. Performance analysis shows that the scheme is at least five orders faster than the Paillier schemes and at least twice as faster than the existing randomisation technique at 128-bit security.

preprint2016arXiv

User Collusion Avoidance Scheme for Privacy-Preserving Decentralized Key-Policy Attribute-Based Encryption -- Full Version

Recent trend towards cloud computing paradigm, smart devices and 4G wireless technologies has enabled seamless data sharing among users. Cloud computing environment is distributed and untrusted, hence data owners have to encrypt their data to enforce data confidentiality. The data confidentiality in a distributed environment can be achieved by using attribute-based encryption technique. Decentralized attribute-based encryption technique is a variant of multiple authority based attribute-based encryption whereby any attribute authority can independently join and leave the system without collaborating with the existing attribute authorities. In this paper, we propose a privacy-preserving decentralized key-policy attribute-based encryption scheme. The scheme preserves the user privacy when users interact with multiple authorities to obtain decryption keys while mitigating the well-known user collusion security vulnerability. We showed that our scheme relies on decisional bilinear Diffie-Hellman standard complexity assumption in contrast to the previous nonstandard complexity assumptions such as $q-$decisional Diffie-Hellman inversion.

preprint2014arXiv

Spontaneous expression classification in the encrypted domain

To date, most facial expression analysis have been based on posed image databases and is carried out without being able to protect the identity of the subjects whose expressions are being recognised. In this paper, we propose and implement a system for classifying facial expressions of images in the encrypted domain based on a Paillier cryptosystem implementation of Fisher Linear Discriminant Analysis and k-nearest neighbour (FLDA + kNN). We present results of experiments carried out on a recently developed natural visible and infrared facial expression (NVIE) database of spontaneous images. To the best of our knowledge, this is the first system that will allow the recog-nition of encrypted spontaneous facial expressions by a remote server on behalf of a client.