Department of Computer Science and Engineering

Indian Institute of Technology Hyderabad (IITH) is a premier institute of science and technology established in 2008. IITH has been consistently ranked in the&nbsp;<a href="https://iith.ac.in/reports/" rel="noopener noreferrer">top 10 institutes</a>&nbsp;in India for Engineering&nbsp;<a href="https://www.nirfindia.org/2021/EngineeringRanking.html" rel="noopener noreferrer" target="_blank">according to NIRF</a>&nbsp;making it one of the most coveted schools for science and technology in the country.

IITH offers&nbsp;<a href="https://iith.ac.in/academics/under-graduate/">undergraduate programs</a>&nbsp;in all the classical engineering disciplines, applied sciences, design, as well as several modern interdisciplinary areas. Students are given a flexibility to explore a broad set of areas, and potentially pursue a minor or double major in a discipline that is not their own. Students who wish to seek a deeper understanding of their own discipline are strongly encouraged to get involved in cutting-edge research with the help of a faculty to mentor them, and earn an Honors in their own field.

The very foundation of IIT Hyderabad is based on&nbsp;<a href="https://iith.ac.in/research/">research and innovation</a>. The vibrant research culture is evident from the number of patents and publications that IITH produces consistently every year. IITH offers&nbsp;<a href="https://iith.ac.in/academics/programmes-offered/">graduate programs</a>&nbsp;at both a masters, and a doctoral level, in several diverse areas. There are separate programs for technology, design, science, and liberal arts. The&nbsp;<a href="https://iith.ac.in/academics/post-graduate/">MTech program</a>&nbsp;is offered in various modes and durations to cater to the ever-growing need of postgraduate level professionals in the country.

IITH encourages and supports&nbsp;<a href="https://iic-iith.netlify.app/index.html" rel="noopener noreferrer" target="_blank">innovation and entrepreneurship</a>&nbsp;at every stage. The minor program in entrepreneurship is open to all students of IITH, and many of the courses are offered by industrialists who are entrepreneurs themselves. There are various&nbsp;<a href="https://iith.ac.in/research/centres-incubators/">innovation centers and incubators</a>&nbsp;that promote entrepreneurship and help mentor young innovators. IITH has been very successful in building tie-ups with leading academic institutions around the globe.

IITH enjoys a very special relationship with Japanese universities and industries that goes beyond academic and research collaborations. In fact, some of the iconic buildings in IITH campus will carry the signature of Japanese architecture.

IITH creates a unique holistic ecosystem for education that offers interactive learning, a very flexible academic structure, cutting-edge research, strong industry collaboration, and entrepreneurship. This is an environment which enables students and faculty to translate their dreams into realities.

IIT Hyderabad

A collection of linear orders on X, say L, is said to realize a partially ordered set (or poset) P= (X, ≼) if, for any two distinct x,y ∈ X, x ≼ y if and only if x ≺Ly, ∀ L∈ L. We call L a realizer of P. The dimension of P, denoted by dim(P) , is the minimum cardinality of a realizer of P. A containment modelMP of a poset P= (X, ≼) maps every x ∈ X to a set Mx such that, for every distinct x,y ∈ X, x ≼ y if and only if Mx⫋ My. We shall be using the collection (Mx)x∈X to identify the containment model MP. A poset P= (X, ≼) is a Containment order of Paths in a Tree (CPT poset), if it admits a containment model MP=(Px)x∈X where every Px is a path of a tree T, which is called the host tree of the model. We show that if a poset P admits a CPT model in a host tree T of maximum degree Δ and radius r, then dim(P)≤lglgΔ+(12+o(1))lglglgΔ+lgr+12lglgr+12lgπ+3. This bound is asymptotically tight up to an additive factor of min(12lglglgΔ,12lglgr). Further, let P(1 , 2 ; n) be the poset consisting of all the 1-element and 2-element subsets of [n] under ‘containment’ relation and let dim(1,2;n) denote its dimension. The proof of our main theorem gives a simple algorithm to construct a realizer for P(1 , 2 ; n) whose cardinality is only an additive factor of at most 32 away from the optimum. © 2020, Springer Nature B.V.

World University root node

Order

Dimension of CPT Posets

The surge in the spread of immensely powerful deep-fakes, pseudo-news, misleading information, lies, propaganda, and false facts, frequently known as fake news, raised questions concerning social media's influence in today's fast-moving society. It is challenging to overcome fake news in traditional centralized systems. Blockchain technology can help in contending with such problems. A novel idea of DeHiDe: Deep Learning-based Hybrid Model to Detect Fake News using Blockchain is presented in this paper. DeHiDe is a blockchain-based framework for legitimate news sharing by filtering out the fake news. It combines the benefit of blockchain with an intelligent deep learning model to reinforce robustness and accuracy in combating fake news's hurdle. © 2021 ACM.

ACM International Conference Proceeding Series

DeHiDe: Deep Learning-based Hybrid Model to Detect FakeNews using Blockchain

Programmable data planes provide exciting opportunities to realize fast, accurate, and data-driven control-loop decisions. Many data plane systems have been proposed for handling network dynamics (e.g., congestion, failures) in near real-time. The core of these systems has packet-processing data-plane algorithms that continuously monitor traffic and respond automatically. Despite their benefits, automatic response to network events lead to increase in potential sources of inputs, and hence, increase in attack surface. This paper takes a step towards securing such systems by (1) identifying possible attacks on recently proposed data-driven data-plane systems; and (2) designing a scalable tool for detecting such attacks at run time. Our approach models plausible expected behavior and uses the model as a reference to check whether the system is under attack. We conduct preliminary experiments to demonstrate the feasibility of our detection methodology. © 2021 ACM.

SPIN 2021 - Proceedings of the 2021 ACM SIGCOMM Workshop on Secure Programmable network INfrastructure

Anomaly Detection in Data Plane Systems using Packet Execution Paths

With the rapid growth of Twitter in recent years, there has been a tremendous increase in the number of tweets generated by users. Twitter allows users to make use of hashtags to facilitate effective categorization and retrieval of tweets. Despite the usefulness of hashtags, a major fraction of tweets do not contain hashtags. Several methods have been proposed to recommend hashtags based on lexical and topical features of tweets. However, semantic features and data sparsity in tweet representation have rarely been addressed by existing methods. In this paper, we propose a novel method for hashtag recommendation that resolves the data sparseness problem by exploiting the most relevant tweet information from external knowledge sources. In addition to lexical features and topical features, the proposed method incorporates the semantic features based on word-embeddings and user influence feature based on users’ influential position. To gain the advantage of various hashtag recommendation methods based on different features, our proposed method aggregates these methods using learning-to-rank and generates top-ranked hashtags. Experimental results show that the proposed method significantly outperforms the current state-of-the-art methods. © 2020, Springer-Verlag London Ltd., part of Springer Nature.

Knowledge and Information Systems

Hashtag recommendation for short social media texts using word-embeddings and external knowledge

Modern-day cyber threats are growing more rapidly than ever before. To effectively defend against them, Anomaly-based Network intrusion detection systems (A-NIDS) must evolve continuously. Traditional machine learning techniques are ineffective in handling sequentially evolving tasks, and Neural Networks (NNs) in particular suffer from Catastrophic Forgetting (CF) of old tasks when trained on new ones. Continual Learning (CL) strategies help to mitigate CF by imposing constraints while training NNs on sequentially evolving data like network traffic. However, applying the CL framework in the design of A-NIDS is not straightforward due to the heavy Class Imbalance (CI) in the network traffic datasets. As a result, the performance of the system is very sensitive to the task execution order. In this work, we propose a CL based A-NIDS by applying sample replay with Class Balancing Reservoir Sampling (CBRS) to mitigate CI in a Class Incremental Setting (CIS). Using the CICIDS-2017 dataset, experiments are conducted by permuting the majority class across the different task execution orders using the proposed CL based A-NIDS. We find that using auxiliary memory with context-aware sample replacing strategies, CF can be reduced to a greater extent, as opposed to data augmentation techniques which may alter the original data distribution and increase training time (with oversampling methods). © 2021 ACM.

Department of Electrical Engineering

On Handling Class Imbalance in Continual Learning based Network Intrusion Detection Systems

New vehicular applications like Augmented Reality (AR), Virtual Reality (VR), and High Definition Map (HD Map) have computational intensive and latency-sensitive traits and require collaboration among nearby vehicles. Computational offloading is used to improve the accuracy and performance of these applications, as it allows computational jobs to be processed on MEC servers at the cell edge. Here, the challenge is how to effectively take offloading decisions at the MEC server by considering wireless transmission delay and computational delay in the presence of time varying channel conditions due to vehicular mobility. In this work, we aim to maximize the number of jobs offloaded to the MEC server under application's deadline constraints while ensuring fairness among vehicles. First, we formulate the computational offloading as an integer linear programming (ILP) problem where both the transmission delay of 5G NR and MEC computational resources are taken into account. Then, we propose an online heuristic for joint computational offloading and resource allocation, OCTANE, that jointly takes 5G NR radio resources and computational resources into consideration while taking offloading decisions. Further, to provide fairness among vehicles, Transport Block Size (TBS) based Medium Access Layer (MAC) strategy is proposed for allocation of TDMA symbols in the 5G NR uplink. Finally, extensive simulations are performed in the NS-3 5G NR module with mobility traces taken from SUMO using OpenStreetMap to evaluate OCTANE and ILP model. Simulation results show that the proposed OCTANE scheme performs better than a state-of-the-art solution and is close to the ILP model in terms of offloading success rate. © 2021 IEEE.

International Symposium on Advanced Networks and Telecommunication Systems, ANTS

OCTANE: A Joint Computation Offloading and Resource Allocation Scheme for MEC Assisted 5G NR Vehicular Networks

Cloud Radio Access Network (C-RAN) disaggregates the functionalities of the base station in a way that some of the radio processing tasks are centralized in a virtualized computer pool of general-purpose processors (GPPs) on a cloud platform. This enables efficient utilization of the computational resources based on the spatio-temporal traffic fluctuations at cell sites. In this paper, we attempt to further reduce the computation resources by C-RAN on the cloud platform. First, we profiled the energy consumed in an OpenAirInterface (OAI) based C-RAN system using the existing Linux CPU frequency scaling governors. Based on the observations, we propose a traffic-aware compute resource tuning (CRT) scheme that reduces the energy consumption of C-RANs. The CRT scheme opportunistically lowers Modulation Coding Scheme (MCS) used while serving users by utilizing all of the available radio resources in every scheduling interval during non-peak hours. This reduction in the MCS helps in reducing energy consumption (due to usage of lower CPU clock frequency in the GPPs of the cloud platform) and fronthaul bandwidth requirements. Another benefit of the CRT scheme is its ability to work with any MAC scheduler. The extensive simulation results show how the CRT outperforms the existing frequency scaling governors in energy consumption while reducing fronthaul bandwidth requirements. © 2021 IEEE.

2021 IEEE Global Communications Conference, GLOBECOM 2021 - Proceedings

Traffic-Aware Compute Resource Tuning for Energy Efficient Cloud RANs

The fifth-generation (5G) mobile technology comes with numerous enhancements to support new use cases and services. Two of these enhancements are Numerology and Bandwidth Part (BWP) in the 5G NR (New Radio) standard. While Numerology gives the flexibility for selecting Transmission Time Interval (TTI) for different types of service classes, the BWP allows a gNB (5G NR Base Station) carrier to be divided into multiple parts. Different BWPs can run different Numerologies simultaneously by selecting higher Numerology for low latency use cases. In this paper, we identified the drawbacks of higher Numerology under high traffic scenario and proposed an optimal bandwidth (BW) segmentation algorithm. The proposed solution reduces the downlink latency by selecting Numerologies and BWP based on the traffic conditions of various services. The simulation results show that the proposed algorithm can effectively adapt to the traffic load and reduce the delay and packet loss rate by up to 36% and 18.5%, respectively compared to static Numerology-based scheduling. © 2021 IEEE.

Traffic Aware Optimal Bandwidth Segmentation for Low Latency Communication in 5G

Prefetching is a technique used to improve system performance by bringing data or instructions in the cache before it is demanded by the core. Several prefetching techniques have been proposed, in both hardware and software, to predict the data to be prefetched with high accuracy and coverage. The memory patterns accessed by applications can be classified as either regular memory access patterns or irregular memory access patterns. Most prefetchers exclusively target either of these patterns by learning from either temporal or spatial correlation among the past data accesses observed. Our proposal focuses on covering all kinds of access patterns which can be predicted by a temporal as well as a spatial prefetcher. Running both kinds of prefetchers in parallel is not a wise design as it leads to unnecessary hardware (storage) overhead for metadata storage of temporal prefetcher. We propose broadly classifying the memory access patterns of applications on the go as regular or irregular, and then using an appropriate prefetcher to issue prefetches for the respective classes. This reduces the metadata requirement in case of temporal prefetcher by 75%. Evaluation of our proposed solution on SPEC CPU 2006 benchmarks achieve a speedup of 23.7% over the no-prefetching baseline, which is a 4% improvement over the state of the art spacial prefetcher BIP, and 13.2% improvement over the temporal prefetcher, Triage. © 2021 ACM.

MAPCP: Memory Access Pattern Classifying Prefetcher

Deep Gaussian Processes (DGPs) are multi-layer, flexible extensions of Gaussian Processes but their training remains challenging. Sparse approximations simplify the training but often require optimization over a large number of inducing inputs and their locations across layers. In this paper, we simplify the training by setting the locations to a fixed subset of data and sampling the inducing inputs from a variational distribution. This reduces the trainable parameters and computation cost without significant performance degradations, as demonstrated by our empirical results on regression problems. Our modifications simplify and stabilize DGP training while making it amenable to sampling schemes for setting the inducing inputs. © 2021 37th Conference on Uncertainty in Artificial Intelligence, UAI 2021. All Rights Reserved.