Determination of suitable resource discovery tool and methodology for high-volume internet of things (IoT)
This paper discusses several issues supporting a knowledge-based methodology for discovery of high volume IoT resources in the simulator NS-3 environment. We found the concept was developed in previous researches, especially based on widely accepted concepts of Q-Learning discovery model. The mode...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | http://eprints.unisza.edu.my/4410/1/FH03-FIK-21-54260.pdf http://eprints.unisza.edu.my/4410/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Universiti Sultan Zainal Abidin |
| Language: | English |
| Summary: | This paper discusses several issues supporting a knowledge-based methodology for discovery of high volume IoT resources in the simulator NS-3 environment. We found the concept was developed in
previous researches, especially based on widely accepted concepts of Q-Learning discovery model. The
model is validated using samples from emulations of tested data in the NS-3 simulator. Proper
simulation in NS-3 based on the different modules such as checkpoint and restore was used to model
and analyse the data. The main feasibility checkpointing concept of simulations in the NS-3 processes
were using Distributed Multi-Threaded Checkpointing (DMTCP) to run on a single machine and
Message Passing Interface (MPI) under distributed machine to speed up the NS-3 model initialization
and execution. As the chosen model to be implemented in this analysis, the Q-learning algorithm
proposal offers a possible solution for addressing evolving IoT environments and configurations. Q-learning is one of the successful techniques available for the exploration of IoT nodes, but context based problems have already been established and simplified as issues of dedicated server
management, IoT object data acquisition issues, and unique application requirements. The findings
empirically support the validation of the Q-Learning model improvement for high-volume IoT resource
discovery cases. The study will contribute to the new model development by providing new insights on
the conceptualization and validation of knowledge-based methodology based on widely accepted
techniques and approaches. |
|---|