Adaptive Serverless FaaS Dataflow Framework for Real-Time IoT Analytics across the Cloud-Edge Continuum

Deshmukh, Pradnya Deepak (2024) Adaptive Serverless FaaS Dataflow Framework for Real-Time IoT Analytics across the Cloud-Edge Continuum. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration Manual) Download (440kB) | Preview

Abstract

Robust cloud-to-edge communication infrastructure is now available due to proliferation of 5G networks and IoT devices. Existing IoT systems, however, do not have efficient mechanisms for workflow orchestration or adaptation to dynamic environments. This research introduces an adaptive serverless framework for real time IoT analytics across the cloud edge continuum and uses intelligent workflow placement to improve the responsiveness of these workloads. The framework realizes a new machine learning based anomaly detection, time series prediction, and reinforcement learning approach to processing location determination. An autoencoder based anomaly detection model that identifies patterns which require special processing, LSTM networks for prediction of processing requirements and a Deep Q network for dynamic placement decision in the edge, hybrid and cloud environments are employed by the system. This framework has been implemented on AWS infrastructure and show results up to 13x superior placement efficiency than traditional rule based approaches with single request latencies of 63.92ms versus 832ms, and consistent performance up to large loads of 80k concurrent users. Its ML-driven decision making improves placement accuracy and adaptation to changing conditions significantly, according to evaluation. Specifically, this research advances the state of the art in the realm of IoT analytics by enabling a flexible orchestration mechanism that can adapt to a number of real world scenarios encountered in smart city, healthcare system, and other IoT applications.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Jaswal, Shivani UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Cloud computing T Technology > TK Electrical engineering. Electronics. Nuclear engineering > Telecommunications > Computer networks > Internet of things
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Ciara O'Brien
Date Deposited:	14 Jul 2025 15:28
Last Modified:	14 Jul 2025 15:28
URI:	https://norma.ncirl.ie/id/eprint/8092

Actions (login required)

View Item