An efficient and secure framework for federated learning enabled IOT systems

dc.contributor.guideMohit Kumar and Renu Dhir
dc.coverage.spatial
dc.creator.researcherMansi Gupta
dc.date.accessioned2025-12-16T10:49:38Z
dc.date.available2025-12-16T10:49:38Z
dc.date.awarded2025
dc.date.completed2025
dc.date.registered2021
dc.description.abstractABSTRACT newlineArtificial Intelligence (AI) has become integral to developing intelligent, data-driven solutions newlinein rich areas of development like medical care, transport, business automation, farming, and newlineindividualized services. The Internet of Things (IoT) increasingly supports such developments, newlineenabling many interconnected devices, such as sensors, wearables, smart meters, and actuators, newlineto collect, process, and share data in real-time. The application of IoT in smart healthcare newlineincludes remote monitoring, traffic control, and driverless vehicles in intelligent transport newlinesystems, precision agriculture, smart city, and smart homes in energy-efficient systems, and newlinesafety, in addition to other applications. newlineConventionally, these applications have been relying on centralized machine learning models, newlinemaking it necessary to aggregate data across different devices into one server to train. This newlinecentralized solution applies to a wide range of datasets. Still, it poses significant challenges when newlinedata privacy, regulatory compliance, and communication overhead are concerned, particularly newlinein an environment where resource-constrained IoT devices generate edge data. Moreover, the newlineheterogeneity and extension of IoT infrastructures lead to devices with varying computing power, newlineinterrupted connectivity, and limited energy, which is an issue for the seamless deployment of newlinecentralized AI systems. newlineFederated Learning (FL) has become one of the potential paradigms to overcome these issues newlinesince it allows decentralized training of models and maintains data privacy without sharing newlineraw data. However, traditional FL frameworks often depend on synchronous communication newlinemodels, suffers from straggler effect, wherein the server awaits the completion of local training newlineby all clients before aggregating updates. This will be unrealistic in an environment that is newlineheterogeneous and an intermittently connected IoT with a device with a variable degree of newlinecomputational capacity and availability. Despite this limitation, synchronous FL is still usefu
dc.description.note
dc.format.accompanyingmaterialNone
dc.format.dimensions
dc.format.extent
dc.identifier.researcherid
dc.identifier.urihttp://hdl.handle.net/10603/681376
dc.languageEnglish
dc.publisher.institutionDepartment of Information Technology
dc.publisher.placeJalandhar
dc.publisher.universityDr B R Ambedkar National Institute of Technology Jalandhar
dc.relation
dc.rightsuniversity
dc.source.universityUniversity
dc.subject.keywordComputer Science
dc.subject.keywordComputer Science Artificial Intelligence
dc.subject.keywordEngineering and Technology
dc.titleAn efficient and secure framework for federated learning enabled IOT systems
dc.title.alternative
dc.type.degreePh.D.

Files

Original bundle

Now showing 1 - 5 of 13
Loading...
Thumbnail Image
Name:
01_title.pdf
Size:
255.5 KB
Format:
Adobe Portable Document Format
Description:
Attached File
Loading...
Thumbnail Image
Name:
02_prelims.pdf
Size:
1.33 MB
Format:
Adobe Portable Document Format
Loading...
Thumbnail Image
Name:
03_content.pdf
Size:
69.08 KB
Format:
Adobe Portable Document Format
Loading...
Thumbnail Image
Name:
04_abstract.pdf
Size:
54.24 KB
Format:
Adobe Portable Document Format
Loading...
Thumbnail Image
Name:
05_chapter1.pdf
Size:
1.11 MB
Format:
Adobe Portable Document Format

License bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
1.79 KB
Format:
Plain Text
Description: