Department of Computer Science and Engineering

We define and construct Deniable Fully Homomorphic Encryption based on the Learning With Errors (LWE) polynomial hardness assumption. Deniable FHE enables storing encrypted data in the cloud to be processed securely without decryption, maintaining deniability of the encrypted data, as well the prevention of vote-buying in electronic voting schemes where encrypted votes can be tallied without decryption. Our constructions achieve compactness independently of the level of deniability- both the size of the public key and the size of the ciphertexts are bounded by a fixed polynomial, independent of the detection probability achieved by the scheme. This is in contrast to all previous constructions of deniable encryption schemes (even without requiring homomorphisms) which are based on polynomial hardness assumptions, originating with the seminal work of Canetti, Dwork, Naor and Ostrovsky (CRYPTO 1997) in which the ciphertext size grows with the inverse of the detection probability. Canetti et al. argued that this dependence “seems inherent”, but our constructions illustrate this is not the case. We note that the Sahai-Waters (STOC 2014) construction of deniable encryption from indistinguishability obfuscation achieves compactness and can be easily modified to achieve deniable FHE as well, but it requires multiple, stronger sub-exponential hardness assumptions, which are furthermore not post-quantum secure. In contrast, our constructions rely only on the LWE polynomial hardness assumption, as currently required for FHE even without deniability. The running time of our encryption algorithm depends on the inverse of the detection probability, thus the scheme falls short of achieving simultaneously compactness, negligible deniability probability and polynomial encryption time. Yet, we believe that achieving compactness is a fundamental step on the way to achieving all properties simultaneously as has been the historical journey for other primitives such as functional encryption. Our constructions support large message spaces, whereas previous constructions were bit by bit, and can be run in online-offline model of encryption, where the bulk of computation is independent of the message and may be performed in an offline pre-processing phase. This results in an efficient online phase whose running time is independent of the detection probability. At the heart of our constructions is a new way to use bootstrapping to obliviously generate FHE ciphertexts so that it supports faking under coercion.

A major issue in the energy-constrained ad hoc wireless networks is to find ways that increase their lifetime. The communication protocols need to be designed such that they are aware of the battery state. They should also consider the presence of varying sizes and models of the nodes' batteries. Traditional MAC protocols of these networks are designed considering neither the state nor the characteristics of the batteries of the nodes. Major contributions of this paper are: (a) a novel distributed heterogeneous Battery Aware MAC (HBAMAC) protocol, that takes benefit of the chemical properties of the batteries and their characteristics, to provide fair node scheduling and increased network and node lifetime through uniform discharge of batteries, (b) theoretical analysis, using discrete time Markov chains, for batteries of the nodes, and (c) a thorough comparative study of our protocol with IEEE 802.11 and DWOP (Distributed Wireless Ordering Protocol) MAC protocols. © Springer-Verlag 2004.

Recently, there has been a tremendous growth in mobile network traffic. Network providers are looking for techniques that selectively offload the mobile data traffic onto WiFi (IEEE 802.11) networks to balance the load and improve network performance. Several architectures based on Proxy Mobile IPv6 (PMIPv6) have been proposed to support seamless data offloading. The demerits of PMIPv6 include lack of flow mobility and single point of failure. There exist architectures that extend PMIPv6 to support flow mobility, but still face the problem of overhead at the gateway and single point of failure. In this paper, we propose Seamless Internetwork Flow Mobility (SIFM), a new architecture that overcomes these drawbacks and provides seamless data offload supporting flow mobility. Both the PMIPv6 and the SIFM architectures have been implemented and evaluated incorporating salient LTE and WiFi network features in the ns-3 simulator. The performance studies validate that seamless mobility can be achieved for clients in both of these architectures. The results show that for the best possible (scenario dependent) offload value, the SIFM architecture shows an improvement of 13.86%, 29.05% and 11.33% whereas the PMIPv6 architecture shows an improvement of 7.96%, 19.52% and 7.83% in terms of delay, packet loss and throughput respectively compared to no offload scenario in each architecture. Further, we also show that the support for flow mobility in the SIFM architecture provides the flexibility to move selective flows to another network. This helps in achieving better performance gain compared to moving all the flows of the user as done in the PMIPv6 architecture.

In this paper, the problem of optimal distribution of measurement data to be processed in minimal time on a hypercube network of sensor driven processors is considered. An analytical model is developed for solving the problem efficiently. Unlike the previous models, this model considers 1) explicitly the setup time which constrains exploiting all the available processors; 2) simultaneous use of links to expedite the communication; 3) partial solution combining time to encompass wider class of related problems. By deriving a lower bound on the amount of data to be received by a processor for efficient distribution, a new technique called fractal hypercube is introduced here to get the optimal solution with fewer processors. An optimal iterative method for hypercubes and a near-optimal recursive method with a refinement are presented for the same with the analysis. The effect of varying the originating processor and the choice of fractal hypercube are discussed with an effective technique called processor isomorphism. This study reveals that always the fractal hypercubes outperform the other two methods, the optimal iterative method for hypercubes and the near-optimal method. © 1998 IEEE.

In Heterogeneous cellular networks (HetNets), co-channel interference is a major concern due to the co-existence of multiple base stations with overlaid regions. Edge users are typically the victims because of high interference exposure. To counter this high interference, picocells communicate with the edge users in protected subframes (PSF). The severity of the problem intensifies in case of hotspot deployment, where picocells cannot provide coverage to the entire hotspot, thus forming a dense ring of macrocell users around picocells. We argue that these macrocell users are also victims, constituting a significant victim user population in hotspot deployment. We propose that, along with the macrocell muting during PSF, picocells should also be operated in cooperative manner with macrocell, and be barred from transmission during some of the subframes for protection of these macrocell victim users. We define a utility function to find the optimal values of PSF density for both macrocell and picocells, which would increase victim user throughput thereby enhancing system fairness. Exhaustive simulations illustrate that the proposed scheme improves victim user throughput significantly, while maintaining the overall system capacity. © 2014 IFIP.

Sanitizable signatures provide several security features which are useful in many scenarios including military and medical applications. Sanitizable signatures allow a semi-trusted party to update some part of the digitally signed document without interacting with the original signer. Such schemes, where the verifier cannot identify whether the message has been sanitized, are said to possess strong transparency. In this paper, we have described the first efficient and provably secure sanitizable signature scheme having strong transparency under the standard model. © 2010 Springer-Verlag.

The rapid growth of multicast applications initiated the need to realise multicasting in Wavelength Division Multiplexing (WDM) networks in an efficient and cost-effective way. However, most of the applications need to support dynamic multicast sessions wherein the membership of the multicast group changes with time as nodes join or leave the session dynamically. In addition, with the advancements in the multimedia and other real-time group applications, the construction of multicast trees satisfying Quality of Service (QoS) requirements of the member nodes is becoming a problem of prime importance. In this paper, we deal with on-line QoS multicast routing in WDM networks. As members leave the sessions, it will be mandatory to rearrange the trees in order to save costs, at the same time accounting for the disruption caused to the ongoing transmissions because of the rearrangement. We try to effectively decrease the disruption by considering only certain portions of the multicast tree and limiting the rearrangement to only those portions. A region's usefulness to the overall tree is judged by a novel concept of Cost-to-Service Ratio (CSR) and a region is selected for rearrangement based on its CSR value. Extensive simulation results reveal that rearrangements based on CSR can reduce the disruption to the member nodes by 20-85% compared to rearrangements based on the best known metric (Quality Factor) in the literature, for the same cost performance. We also demonstrate the cost-effectiveness of the trees generated by our on-line algorithm compared to that of static algorithm. © 2006 Elsevier B.V. All rights reserved.

This paper describes ways to improve prosody modeling in syllable-based concatenative speech synthesis systems for two Indian languages, namely Hindi and Tamil, within the unit selection paradigm. The syllable is a larger unit than the diphone and contains most of the coarticulation information. Although syllable-based synthesis is quite intelligible compared to diphone based systems, naturalness especially in terms of prosody, requires additional processing. Since the synthesizer is built using a cluster unit framework, a hybrid approach, where a combination of both rule based and statistical models are proposed to model prosody of syllable like units better. It is further observed that prediction of phrase boundaries is crucial, particularly because Indian languages are replete with polysyllabic words. CART based phrase modeling for Hindi and Tamil are discussed. Perceptual experiments show a significant improvement in the MOS for both Hindi and Tamil synthesizers.

HMM based speech synthesis (HTS) is a state-of-the art approach to text-to-speech synthesis. Segmentation of the training data is essential for building any text-to-speech system. Most conventional text-to-speech systems use phones as the basic unit of synthesis and use a speech recogniser to automatically segment the data at the phone level. As Indian languages are low resource languages, accurate transcriptions are difficult to obtain owing to paucity of data. Manual labeling at the phone level is not only laborious but also inaccurate. HMM based flat start segmentation doesn't work well at the sentence level. In this paper we propose an event driven approach to obtain better phone boundaries. Syllable-like events are detected in the speech signal and matched with syllabified transcription of the text. The syllables are converted to phoneme sequences and Baum-Welch embedded re-estimation is restricted to the syllable-level. Subjective evaluations indicate that the proposed system has a lower word error rate compared to that of a conventional system that uses flat start for obtaining phone boundaries. © 2014 IEEE.

The focus of this paper is two-fold: (a) to develop a knowledge-based robust syllable segmentation algorithm and (b) to establish the importance of accurate segmentation in both the training and testing phases of a speech recognition system. A robust segmentation algorithm for segmenting the speech signal into syllables is first developed. This uses a non-statistical technique that is based on group delay(GD) segmentation and Vowel Onset point(VOP) detection. The transcription corresponding to the utterance is syllabified using rules. This produces an annotation for the train data. The annotated train data is then used to train a syllable-based speech recognition system. The test signal is also segmented using the proposed algorithm. This segmentation information is then incorporated into the linguistic search space to reduce both computational complexity and word error rate(WER). WER's of 4.4% and 21.2% are reported on the TIMIT and NTIMIT databases respectively. ©2010 IEEE.