Department of Computer Science and Information Systems

Permanent URI for this collectionhttp://localhost:4000/handle/123456789/1928

Browse

Now showing 1 - 20 of 1099

STATE Distributed algorithm for uniform circle formation by multiple mobile robots
(Springer, 2016-08) Mohan, Sudeept; Gautam, Avinash
Multi-robot pattern formation has a wide range of applications, such as inspection of hazardous regions, parallel and simultaneous transportation of load, area exploration, etc. This problem has been investigated both from the scientific and engineering perspectives. There is a whole body of experimental work on robot formations, and at the same, there is no dearth of theoretical research addressing the same problem. Several assumptions considered in these theoretical studies are overly simplified with an understanding that they will somehow be reasonably approximated. Although these theoretical research works are sound and complete, they do not show how an actual implementation compares with the idealized scenario. A new practical model is suggested in this paper for geometric pattern formation. This model uses approximate solutions to some of the assumptions considered in theoretical research works. A novel algorithm, STATE, is proposed and is shown to perform better than Défago and Konagaya’s algorithm for uniform circle formation. Both the algorithms are implemented on a real multi-robot test bed. A new framework for inter-robot communication is developed. It supports seamless asynchronous and non-blocking robot-to-computer, and robot-to-robot communication.
FAST Synchronous Frontier Allocation for Scalable Online Multi-Robot Terrain Coverage
(Springer, 2016-09) Mohan, Sudeept; Gautam, Avinash
We propose Frontier Allocation Synchronized by Token passing (FAST), a distributed algorithm for online terrain coverage using multiple mobile robots, ensuring mutually exclusive selection of frontier cells. Many existing approaches cover the terrain in an irregular fashion, without considering the usability of the already covered region. For instance, in the task of floor cleaning in an office building, these approaches do not guarantee the cleanliness of large unbroken areas until a majority of the task is complete. FAST on the other hand, incrementally traverses the terrain generating structured trajectories for each robot. Following a structured trajectory for coverage path planning is proven to be a very powerful approach in literature. This renders large portions of the terrain usable even before the completion of the coverage task. The novel map representation techniques used in FAST render it scalable to large terrains, without affecting the volume of communication among robots. Moreover, the distributed nature of FAST allows incorporation of fault-tolerance mechanisms.
Energy efficient real-time scheduling algorithm for mixed task set on multi-core processors
(Inder Science, 2017) Mohan, Sudeept
Energy optimisation is gaining greater significance in a wide range of systems from mobile devices to datacentres. Specifically, in battery powered real-time embedded systems where tasks are executed under hard timing constraints, energy optimisation poses a big challenge. This paper focuses on dynamic energy optimisation using a well-established technique namely dynamic voltage and frequency scaling (DVFS). This work presents a real-time scheduling algorithm that uses DVFS on mixed task system containing periodic as well as aperiodic tasks on homogeneous multi-core processor. The proposed algorithm guarantees periodic task deadlines and offers minimum aperiodic task response times. Simulation analysis shows that the proposed scheme saves more energy as compared to cycle conserving, static FVS and non-DVFS scheduling algorithms. Further, it does not result in any response time degradation of aperiodic tasks as compared to other algorithms.
Cluster, Allocate, Cover: An Efficient Approach for Multi-robot Coverage
(IEEE, 2015) Mohan, Sudeept; Gautam, Avinash
This article presents an algorithm for online multirobot coverage that proceeds with minimal knowledge of the already explored region and the frontier cells. It creates clusters of frontier cells which are designated to robots using an optimal assignment scheme. Coverage is then performed using a novel path planning technique. Many approaches that use clustering for multi-robot coverage do not specify strict time criteria for re-clustering. Moreover, the motion plans they use result in redundant coverage. To overcome these limitations, an appropriate motion plan for the robots is chosen based on the context of already covered frontiers. Dispersion of robots is vital for efficient coverage and is an emergent behavior in our approach. The efficacy of the proposed approach is tested in simulation and on a multi-robot test-bed. The algorithm performs better than some state of the art approaches.
A Simulation Tool for Energy Efficient Real Time Scheduling and Analysis
(ECRTS, 2015-07)
This paper proposes an automation tool-STREAM, for simulation, testing and analysis of energy aware real time scheduling algorithms for periodic as well as mixed task sets on multi-core processor. The key features of STREAM include implementation of scheduling algorithms, synthetic task set generation, modules for doing performance analysis and generation of execution traces. The simulator design is made simple, understandable and flexible such that addition of new algorithms or modification to existing algorithms can be done with minimum efforts. Testing of modules is carried out on randomly generated task sets. STREAM facilitates the comparative performance analysis of different scheduling algorithms on uniprocessor, multiprocessor or multi-core processor platforms. A graph plotter is provided to visualize the performance analysis of different scheduling algorithms.
Dynamicvoltage and frequency scaling scheduling algorithm for mixed task set
(IEEE, 2013) Mohan, Sudeept
Dynamic voltage and frequency scaling (DVFS) is one of the popular energy conservation techniques for battery operated real-time and embedded systems. This paper presents EEDVFS, an energy efficient DVFS scheduling algorithm for mixed task set. EEDVFS is a variant of Earliest Deadline First (EDF) based Deferrable Server (DS). Experimental results show that EEDVFS reduces energy consumption without compromising on periodic task deadlines and aperiodic task responsiveness. The results of EEDVFS are compared with non-DVFS EDF version of DS upon various performance metrics such as energy consumption, average response time, latency, preemption count and number of decision points. EEDVFS offers 54% energy saving in comparison with the non-DVFS EDF based DS algorithm.
A review of research in multi-robot systems
(IEEE, 2012) Gautam, Avinash; Mohan, Sudeept
Formally, a collection of two or more autonomous mobile robots working together are termed as teams or societies of mobile robots. In multi robot systems simple robots are allowed to coordinate with each other to achieve some well defined goals. In these kinds of systems robots are far less capable as an entity, but the real power lies in cooperation of multiple robots. The simplicity of multi-robots have produced a potentially wide set of applications such as military missions (battlefield surveillance), searching for survivors in disaster hit areas, parallel and simultaneous transportation of vehicles, and delivery of payloads. Although the research on multi-robot systems has attracted considerable attention worldwide in the past decade, the research in this area is still in its infancy. This paper surveys various interaction techniques in multi robot systems which are important with respect to goal attainment and task completion.
A practical framework for uniform circle formation by multiple mobile robots
(IEEE, 2012) Gautam, Avinash; Mohan, Sudeept; Misra, Janardan Prasad
This paper gives a software framework for positioning multiple mobile robots in a circular formation. The robots are initially arbitrarily placed on a 2D plane. The definition of the circle formed by these robots is governed by two strict rules (a) all robots should be positioned on the circle circumference (b) all robots should be uniformly placed on circle circumference. The software framework proposed in this paper utilizes two fundamental design patterns the decorator and the observer. Each robot is assigned a unique identity, which is used for conflict resolution in various situations. The model treats the environment as a first class entity providing partial support to the robots for circle formation. All robots are situated in this environment and when a robot is first introduced its position is known in terms of its x and y coordinates. The approach suggested in this paper is a leader-follower approach wherein the environment that determines the leader and then the leader determines the positions on the circle circumference for the follower robots, such that, the total distance travelled by each follower is minimum and therefore the distance travelled by all the robots to form a circle is minimized.
A token passing approach for circle formation by multiple mobile robots
(IEEE, 2013-08) Gautam, Avinash; Mohan, Sudeept; Shekhawat, Virendra Singh
This paper proposes a weakly centralized distributed approach for positioning multiple mobile robots in a circular formation based on token passing. The problem of the circle formation with multiple robots which are arbitrarily placed on a 2D plane requires all robots to be uniformly positioned (i.e., at an equal angular distance of 2π/N, where N = number of robots) on the circle circumference. The suggested approach is a leader-follower approach wherein it is the leader robot which computes the uniform positions on the circle circumference for all the follower robots. The problem of circle formation is divided into two sub-problems (a) leader selection and (b) finding enviable positions for the follower robots from the set of uniform positions computed by the leader robot. Both these problems are solved by token passing so as to reduce communication load on both the leader and the follower robots. The introduction of token passing makes it a weakly centralized framework thereby reducing the burden on the leader robot.
Energy aware real time scheduling algorithm for mixed task set
(IEEE, 2013-09) Mohan, Sudeept
Energy consumption is one of the major limiting factors of battery operated real-time systems. Optimizing energy consumption without affecting performance and schedulability is the major topic to be researched. In this paper, an energy aware real time scheduling algorithm is proposed for a system with mixed task set consisting of both periodic and aperiodic tasks. Dynamic energy reduction techniques like Dynamic Voltage and Frequency Scaling (DVFS) is used for energy optimization without affecting the responsiveness of aperiodic tasks. Performance of the proposed algorithm is compared with non-DVS algorithm. Experimental evaluation reveals that the proposed algorithm saves 54.44% of energy in comparison with non-DVS algorithms. It achieves this with no degradation in responsiveness of the aperiodic tasks.
Design and Development of a Real Time Scheduling Algorithm for Mixed Task Set on Multi-core Processors
(IEEE, 2014-08) Mohan, Sudeept
This paper presents a real time scheduling algorithm for mixed task set on homogeneous multi-core platform. Periodic tasks are scheduled using Partitioned Earliest Deadline First (P-EDF) technique. Aperiodic tasks are assigned globally to different processor cores and scheduled using Total Bandwidth Server (TBS) on each core. In the proposed algorithm, the excess processing capacity of the cores left unused by the periodic tasks can be utilized by assigning aperiodic task to each core. This improves the overall utilization of individual core. Work conserving nature of global assignment reduces response time of aperiodic task. The proposed algorithm is implemented using java based simulator and tested on large number of synthetic test data. Results show improvement in utilization of individual processing core and improvement in response time of aperiodic tasks.
A Comparison of Machine Learning Attributes for Detecting Malicious Websites
(IEEE, 2019-01) Goyal, Navneet
The number of Malicious Websites has increased manifold in the past few years. As on start of year 2018, 1 in every 13 URL was malicious, amounting to 7.8% URLs identified as malicious [1]. These figures have increased by 2.8%, thereby showing an increasing trend of attack vectors through Malicious Websites. These statistics clearly highlight the need to detect Malicious Websites on the Internet. Many research works have suggested Machine Learning techniques to detect Malicious Websites. Research has also been done to compare Machine Learning algorithms for their detection. However, the aspect of attribute selection for detecting Malicious Websites using Machine Learning has not been delved in detail. In Machine Learning techniques, attribute selection outweighs the importance of any other aspect in the process. Thus, there is a need to compare and analyze the various attributes that can help find Malicious Websites faster and better. This paper is focused to address this research gap, so that, fewer and optimal attributes can do a better job
MalCrawler: A Crawler for Seeking and Crawling Malicious Websites
(ACM Digital Library, 2017-01) Goyal, Navneet
Over the years, internet has become the major source of security threat to computer systems. With the number of people browsing internet increasing exponentially in the last couple of years, browser based attacks have become the preferred means of infecting a computer system. These browser based attacks, known as 'Drive-by Download' attacks, inject malicious JavaScript from the server hosting the malicious web application to the browser. Since, the numbers of malicious websites launching such attacks have increased in the past few years; it has become critical to detect them. Typically, search for malicious web pages involves three steps- crawling URLs on the internet, using fast analysis filters to reject benign pages, and then running complex but slow detailed analysis using Honey Clients on the filtered list. While effective, these techniques consume substantial time and computing resources. This limitation can be overcome by designing a crawler which can seek more malicious sites than benign sites, thus, increasing the "toxicity" of the URLs collected in the first step. In this paper, we propose a focused web crawler, named "MalCrawler", which has been designed to crawl and search malicious websites efficiently. This crawler, when compared to a generic crawler, will not only seek more malicious sites than benign sites, but will also handle cloaking, entanglement and AJAX content in malicious sites. MalCrawler, designed, developed and tested, as part of the scope of this paper, proved to be more efficient than generic crawlers.
Understanding and Mitigating Threats from Android Hybrid Apps Using Machine Learning
(IEEE, 2020) Goyal, Navneet
he Android platform has emerged as the most popular computing platform that has more than 2.5 billion devices [1] working across the globe. These devices include not only mobiles and tablets, but even Android Auto modules in cars, various Android versions running on Televisions, watches and host of other smart devices. What makes things more challenging and interesting for the Android Developers and security experts is the fact that various versions of Android Operating System, from Android 2.3.3 (Ginger Bread) to Android 11.0 coexist in this ecosystem. This paper discusses threats that emanate from Hybrid Android Apps. These Hybrid Apps use WebView Component for handling web content within Android Apps. WebView allows HTML and JavaScript to run and render webpages inside Apps, thereby allowing them to download content from Web Servers on the Internet. It is used by several popular Apps, like Facebook, Twitter, Instagram, etc. WebView even allows JavaScript code to call Android code for completing various tasks. While this feature gives tremendous capability to create interactive Hybrid Apps, however, it also opens a route for malicious content to infect the Android Platform using targeted JavaScript based malwares. Any malicious JavaScript, from untrusted or even from trusted source, can thus find its way to exploit this unique linkage with Android Platform. In this paper we analyze Android Web View's security vulnerabilities, access authorization, kind of attacks that it can encounter, and mechanisms to prevent these attacks. To do so, we have developed two Android Apps, viz., "WebView Tool" and "WebView Monitor". Our analysis and detection mechanisms are based on Machine Learning techniques.
A Rapid Prototyping Approach for High Performance Density-Based Clustering
(IEEE, 2019-10) Goyal, Navneet; Goyal, Poonam
Big Data has significantly increased the dependence of data analytics community on High Performance Computing (HPC) systems. However, efficiently programming an HPC system is still a tedious task requiring specialized skills in parallelization and the use of platform-specific languages as well as mechanisms. We present a framework for quickly prototyping new/existing density-based clustering algorithms while obtaining low running times and high speedups via automatic parallelization. The user is required only to specify the sequential algorithm in a Domain Specific Language (DSL) for clustering at a very high level of abstraction. The parallelizing compiler for the DSL does the rest to leverage distributed systems - in particular, typical scale-out clusters made of commodity hardware. Our approach is based on recurring, parallelizable programming patterns known as Kernels, which are identified and parallelized by the compiler. We demonstrate the ease of programming and scalable performance for DBSCAN, SNN, and RECOME algorithms. We also establish that the proposed approach can achieve performance comparable to state-of-the-art manually parallelized implementations while requiring minimal programming effort that is several orders of magnitude smaller than those required on other parallel platforms like MPI/Spark.
μDBSCAN: An Exact Scalable DBSCAN Algorithm for Big Data Exploiting Spatial Locality
(IEEE, 2019) Goyal, Navneet; Goyal, Poonam; Challa, Jagat Sesh
DBSCAN is one of the most popular and effective clustering algorithms that is capable of identifying arbitrary-shaped clusters and noise efficiently. However, its super-linear complexity makes it infeasible for applications involving clustering of Big Data. A major portion of the computation time of DBSCAN is taken up by the neighborhood queries, which becomes a bottleneck to its performance. We address this issue in our proposed micro-cluster based DBSCAN algorithm, μDBSCAN, which identifies core-points even without performing neighbourhood queries and becomes instrumental in reducing the run-time of the algorithm. It also significantly reduces the computation time per neighbourhood query while producing exact DBSCAN clusters. Moreover, the micro-cluster based solution makes it scalable for high dimensional data. We also propose a highly scalable distributed implementation of μDBSCAN, μDBSCAN-D, to exploit a commodity cluster infrastructure. Experimental results demonstrate tremendous improvements in performance of our proposed algorithms as compared to their respective state-of-the-art solutions for various standard datasets. μDBSCAN-D is an exact parallel solution for DBSCAN which is capable of processing massive amounts of data efficiently (1 billion data points in 41 minutes on a 32 node cluster), while producing a clustering that is same as that of traditional DBSCAN.
AnySC: Anytime Set-wise Classification of Variable Speed Data Streams
(IEEE, 2018-12) Goyal, Navneet; Goyal, Poonam; Challa, Jagat Sesh
Classification of data streams has gained a lot of popularity in recent years owing to its multiple applications. In certain applications like community detection from text feeds, website fingerprinting attack, etc., it is more meaningful to associate class labels with groups of objects rather than the individual objects. This kind of classification problem is known as the set-wise classification problem. The few algorithms available in literature for this problem are budget algorithms, i.e. they are designed to process fixed maximum stream speed, and are not capable of handling variable and high speed streams. We present ANYSC which is the first anytime set-wise classification algorithm for data streams. ANYSC handles variable inter-arrival rate of objects in the stream and performs classification of test entities within any available time allowance, using a proposed data structure referred to as CProf-forest. The experimental results show that ANYSC brings in the features of an anytime algorithm and outperforms the existing approaches.
Optical Character Recognition for Sanskrit Using Convolution Neural Networks
(IEEE, 2018) Goyal, Navneet
Ancient Sanskrit manuscripts are a rich source of knowledge about Science, Mathematics, Hindu mythology, Indian civilization, and culture. It therefore becomes critical that access to these manuscripts is made easy, to share this knowledge with the world and to facilitate further research on this Ancient literature. In this paper, we propose a Convolutional Neural Network (CNN) based Optical Character Recognition system (OCR) which accurately digitizes Ancient Sanskrit manuscripts (Devanagari Script) that are not necessarily in good condition. We use an image segmentation algorithm for calculating pixel intensities to identify letters in the image. The OCR considers typical compound characters (half letter combinations) as separate classes in order to improve the segmentation accuracy. The novelty of the OCR is its robustness to image quality, image contrast, font style and font size, which makes it an ideal choice for digitizing soiled and poorly maintained Sanskrit manuscripts.
Pattern-Based Automatic Parallelization of Representative-Based Clustering Algorithms
(IEEE, 2018) Goyal, Poonam; Goyal, Navneet
Ease of programming and optimal parallel performance have historically been on the opposite side of a tradeoff, forcing the user to choose. With the advent of the Big Data era and rapid evolution of sequential algorithms, the data analytics community can no longer afford the tradeoff. We observed that several clustering algorithms often share common traits - particularly, algorithms belonging to same class of clustering exhibit significant overlap in processing steps. Here, we present our observation on domain patterns in Representative-based clustering algorithms and how they manifest as clearly identifiable programming patterns when mapped to a Domain Specific Language (DSL). We have integrated the signatures of these patterns in the DSL compiler for parallelism identification and automatic parallel code generation. Our experiments on different state-of-the-art parallelization frameworks shows that our system is able to achieve near-optimal speedup while requiring a fraction of the programming effort, making it an ideal choice for the data analytics community.
Linguistic Patterns and Cross Modality-based Image Retrieval for Complex Queries
(ACM Digital Library, 2018) Goyal, Navneet; Goyal, Poonam
With the rising prevalence of social media, coupled with the ease of sharing images, people with specific needs and applications such as known item search, multimedia question answering, etc., have started searching for visual content, which is expressed in terms of complex queries. A complex query consists of multiple concepts and their attributes are arranged to convey semantics. It is less effective to answer such queries by simply appending the search results gathered from individual or subsets of concepts present in the query. In this paper, we propose to exploit the query constituents and relationships among them. The proposed approach finds image-query relevance by integrating three models - the linguistic pattern-based textual model, the visual model, and the cross modality model. We extract linguistic patterns from complex queries, gather their related crawled images, and assign relevance scores to images in the corpus. The relevance scores are then used to rank the images. We experiment on more than 140k images and compare the NDCG@n scores with the state-of-the-art image ranking methods for complex queries. Also, ranking of images obtained by our approach outperforms than that of obtained by a popular search engine.