Offloading Scheme for Speech Recognition Applications

Abimbola, Praise Olamide (2023) Offloading Scheme for Speech Recognition Applications. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration manual) Download (5MB) | Preview

Abstract

The adoption of Smart Mobile Devices (SMD) is increasing rapidly. Challenges such as limited battery life, storage capacity, bandwidth, device heterogeneity, and security are hindering the use of SMDs for computation-intensive tasks. To overcome these challenges, Mobile Cloud Computing (MCC) provides a solution by offloading such tasks to the cloud. It involves transferring data processing and storage from mobile devices to the cloud infrastructure. The objective of this research is to provide a solution to the management of SMD resources by developing an offloading framework for a real-time speech recognition mobile application on a SMD.

To achieve this, a real-time speech to text mobile application has been developed. This mobile application require considerable computational resources, which can lead to performance and power consumption issues in the SMD. An off loading decision engine has been developed, the decision to offload is made based on the battery life and the network strength of the SMD. If the battery life of the SMD is below 20%, and the network is stable, the application will be offloaded to the cloud using Google Speech API. If the battery life of the SMD is above 20%, and the network is unstable, the application will run locally using flutter speech-to-text package. The results shows that with a strong network connection and low battery level, offloading resource and computation-intensive tasks to the cloud results in lower CPU and memory usage compared to local processing. However, in situations with poor network connectivity or a good battery life, local processing becomes the preferred choice.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Sahni, Vikas UNSPECIFIED
Subjects:	T Technology > T Technology (General) > Information Technology > Cloud computing P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing H Social Sciences > HG Finance > Fintech T Technology > T Technology (General) > Information Technology > Fintech Q Science > QA Mathematics > Computer software > Mobile Phone Applications T Technology > T Technology (General) > Information Technology > Computer software > Mobile Phone Applications
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Tamara Malone
Date Deposited:	10 Aug 2024 12:37
Last Modified:	10 Aug 2024 12:37
URI:	https://norma.ncirl.ie/id/eprint/7040

Actions (login required)

View Item