Scribd
Upload
16 views12 pages

Traffic Offloading in IoT Using ML

The document presents a Response-aware Traffic Offloading Scheme (RTOS) utilizing regression machine learning to address unpredictable traffic patterns and service quality in large-scale IoT environments. It introduces an adaptive multivariate spline regression model to classify traffic, demonstrating improved processing ratios and reduced delays through simulation. Future directions include enhancing RTOS for real-time adaptation and exploring more efficient machine learning models for traffic prediction.

Uploaded by

royalgaruda96
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
16 views12 pages

Traffic Offloading in IoT Using ML

The document presents a Response-aware Traffic Offloading Scheme (RTOS) utilizing regression machine learning to address unpredictable traffic patterns and service quality in large-scale IoT environments. It introduces an adaptive multivariate spline regression model to classify traffic, demonstrating improved processing ratios and reduced delays through simulation. Future directions include enhancing RTOS for real-time adaptation and exploring more efficient machine learning models for traffic prediction.

Uploaded by

royalgaruda96
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 12

Internet

&
Intranet

Traffic Offloading Scheme using


Regression Machine Learning
Members:
Avahan Tamrakar
Javed Ansari
Kripesh Nihure
content Overview
Problems addressed
Contribution of the paper
Experimental methodology and analysis
Outcome
Future directions provided
overview
Title: A Response-aware Traffic Offloading Scheme using Regression
Machine Learning for User-Centric Large-Scale Internet of Things

Authors: Gunasekaran Manogaran, Gautam Srivastava, Bala Anand Muthu, S. Baskar,


P. Mohamed Shakeel, Ching-Hsien Hsu, Ali Kashif Bashir, Priyan M. Kumar

Publisher: Institute of Electrical and Electronics Engineers (IEEE)

Published in: IEEE Internet of Things Journal, Vol. 8 (5), Pages 3360-3368
Major Problems Addressed

Handling unpredictable traffic patterns Ensuring reliable traffic offloading to


and user density in large-scale IoT manage user request traffic and service
environments. responses effectively.

Addressing the challenge of service quality Smoothing IoT traffic in a view to improve
and resource allocation failures due to the performance and reliability of resource
varying traffic demands. allocation and sharing systems
Contribution of the Paper

Proposes a Response-aware Traffic Offloading Scheme (RTOS) to manage delay-sensitive

user requests.

Introduces an adaptive multivariate spline regression machine learning model for classifying

traffic and reducing failure rates.

Integrates Cyber-Physical Systems (CPS) with IoT-Cloud architecture for efficient decision-

making in offloading methods.

Provides a simulation-based evaluation of RTOS, demonstrating improvements in request

processing ratio, response time, and delay reduction.


Methodology

Techniques Used: Simulation Setup:

Multivariate spline regression learning Conducted using Contiki Cooja simulator.

Simulated environment with 800 users, 8 cloud


model for traffic classification.
servers (each with 2 TB storage and 8 GB
Independent and shared offloading
memory), and 6 CPS nodes.
processes for traffic management.
32 access points for communication between
CPS-assisted decision-making using
cloud, CPS, and IoT devices.
knowledge base and event logs.
Methodology

Performance Metrics Analyzed: Comparative Analysis Against Existing

Processing time Methods:

Processing ratio DTOS-LBBD (Dynamic Task Offloading

and Scheduling)
Average delay
APS-QL (Adaptive Packet Scheduling
Response time
using Q-learning)

ATO (Adaptive Task Offloading)


Analysis of processing time
Analysis of processing %
Outcome

RTOS (Response-aware Traffic Offloading Scheme) demonstrated a


higher processing ratio and reduced processing time and delay compared
to existing models.
Improved response time, reducing failures due to request overload.
Successfully classified and handled traffic based on user density and
service demand.
Future
Directions

Enhancing RTOS for real-time adaptation in dynamic IoT environments.

Exploring more efficient machine learning models for traffic prediction.

Extending the framework to handle


thank you

You might also like