Sudipta Saha

Sudipta Saha

Auto-generated profile, based on publicly available information

Sudipta Saha

School of Electrical and Computer Sciences, Indian Institute of Technology Bhubaneswar, Bhubaneswar, India

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
28 September 2022 in Robotics and Autonomous Systems
Verified
Journal Article
Closed access
PMPrabhat Mahato
Prabhat Mahato
SSSudipta Saha
Sudipta Saha
CSChayan Sarkar
Chayan Sarkar
SHShaghil
Shaghil

17 July 2023 in IEEE Systems Journal
Verified
Journal Article
Closed access
JDJagnyashini Debadarshini
Jagnyashini Debadarshini
MTMadhav Tummala
Madhav Tummala
SSSudipta Saha
Sudipta Saha
OLOlaf Landsiedel
Olaf Landsiedel
MCMun Choon Chan
Mun Choon Chan
Many-to-many/all-to-all sharing of data is one of the common requirements in any distributed/decentralized system. In low-power wireless distributed systems composed of highly resource-constrained devices, a many-to-many data-sharing strategy needs to be not only efficient but also sufficiently simple in nature. Advancements in the field of synchronous transmission (ST) have proliferated a number of efficient data-sharing strategies for low-power systems. In this work, we first identify the fundamental and simple possible approaches to accomplish the goal. Through extensive simulation-based comparison among them, we demonstrate the superiority of the approach adopted by a strategy called MiniCast where the desired task is achieved through pipelined execution of the one-to-many floods from the designated source nodes. However, through further in-depth study, we identify certain crucial flaws in MiniCast, which are mostly caused by its substandard realization through radio-event-triggered architecture. To circumvent the problems, we propose a new strategy TimeCast where the radio-event-triggered architecture of MiniCast is replaced by a time-triggered architecture. Through rigorous experimental studies in IoT-testbeds, we demonstrate that TimeCast substantially improves the robustness, and thus, due to lesser failures, it achieves all-to-all data-sharing up to 57.1% faster while consuming up to 63.6% lesser radio- on time compared to MiniCast.

1 January 2018
Verified
Conference Paper
Closed access
SRShinjini Ray
Shinjini Ray
STSayan Tapadar
Sayan Tapadar
SCSuhrid Krishna Chatterjee
Suhrid Krishna Chatterjee
RKRobin Karlose
Robin Karlose
SSSudipta Saha
Sudipta Saha
HSHimadri Nath Saha
Himadri Nath Saha
Ubiquitous objects are getting “smarter” and more “connected”, every day. With this ever-growing Internet of Things, every object can now be uniquely identified and made to communicate with each other. This approach has been applied to dustbins too, to monitor garbage collection, throwing light on numerous valuable insights. Our project too employs a similar approach, to not only monitor garbage collection but also optimize it, using machine learning. The method of unsupervised learning we utilize is K Means Clustering, widely used in data mining and analytics. Our physical device uses an ultrasonic sensor to be aware of a dustbin's current content level. If the level reaches or exceeds a threshold percentage of the total capacity of the dustbin, it informs our servers, via an online application programming interface (API) developed for this purpose. The API also stores related data - fill time, cleanup time, and location, to name a few. This dynamic dataset generated is analyzed by our algorithm, to determine the times of the day, when a regular cleanup should be performed, such that the dustbins are clean, for the maximum possible portion of the day. The algorithm also shows the locations, where another dustbin should be installed, for further optimization. This is found out by inspecting each cluster individually and scanning out - items which are the furthest away from its closest centroid; and multiple items related to the same dustbin. In either case, a new dustbin installation is advised at such locations. Data henceforth generated revealed that the installation has had a positive effect on the optimization.

9 November 2023 in Ad Hoc Networks
Verified
Journal Article
Closed access
JDJagnyashini Debadarshini
Jagnyashini Debadarshini
RVRejeti Megha Vardhan
Rejeti Megha Vardhan
SSSudipta Saha
Sudipta Saha
Chandrashekhar Narayan BhendeChandrashekhar Narayan Bhende
Chandrashekhar Narayan Bhende

12 August 2024 in ACM Transactions on Internet of Things
Verified
Journal Article
Open access
CSChandra Shekhar
Chandra Shekhar
JDJagnyashini Debadarshini
Jagnyashini Debadarshini
SSSudipta Saha
Sudipta Saha
Detection of vehicles and their classification is a significant component of wide-area monitoring and surveillance, as well as intelligent- transportation . Existing solutions tend to employ heavy-weight infrastructure and costly equipment, as well as largely depend on constant support from the cloud through round-the-clock internet connectivity and uninterrupted power supply. Moreover, existing works mainly concentrate on localized measurement and do not discuss their efficient integration to address the problem over a wide area. For practical use in an outdoor environment, apart from being technically sound and accurate, a solution also needs to be cost-effective , lightweight , easy to install , flexible , low overhead , and easily maintainable , as well as self-sufficient as much as possible. However, fulfilling all these goals together is a challenging task. In this work, we propose an IoT-assisted strategy, LiVeR , to accomplish it. For self-sufficient on-the-fly classification in resource-constrained low-power IoT devices, LiVeR minimizes not only the computational requirements but also the energy consumption, which enables sustained operation in a hostile outdoor environment for a considerably long time solely based on battery power. Through extensive studies based on outdoor measurement and trace-based simulation on empirical data, we demonstrate that LiVeR classifies vehicles of small, medium, and large size with an accuracy of 91.3% up to 98.8%, 92.3% up to 98.5%, and 93.8% up to 98.8%, respectively, for single-lane traffic. We also demonstrate that LiVeR spends only about one-third of the number of RF packets to achieve vehicle detection and classification compared to the state-of-the-art RF-based solution, considerably extending the lifetime of the system.

12 March 2024 in IEEE Transactions on Network and Service Management
Verified
Journal Article
Closed access
JDJagnyashini Debadarshini
Jagnyashini Debadarshini
MKManish H. Kausik
Manish H. Kausik
SSSudipta Saha
Sudipta Saha
A significant part of an IoT-based smart-system is comprised of a large number of low-cost devices that interact with one another based on low-power wireless communication. Energy-efficient and fast sharing of data plays a crucial role in such massive distributed systems. The topology of the underlying network highly influences the efficiency of a data-sharing process. However, since topological information becomes available only after the deployment of a system, it is hard to design a strategy paying proper attention to this important issue. In this work, we explore how many-to-many data-sharing, which is the most complex and highly important pattern of sharing data, can benefit from the availability of network topology information. We also deal with how the necessary topology information can be obtained in run-time. In particular, we design a Synchronous-Transmission (ST)-based many-to-many data-sharing strategy that can dynamically modulate itself in real-time as per the topology of the underlying network. Through extensive simulation as well as experiments over publicly available IoT-testbeds we demonstrate that the proposed strategy can accomplish all-to-all data-sharing in IoT up to 42.5% faster and consume 43.3% less radio-on time compared to the existing state-of-the-art strategies.