Dey, Prasanjit and Kumar, Chandan and Mitra, Mitrabarun and Mishra, Richa and Chaulya, S.K. and Prasad, G.M. and Mandal, S.K. and Banerjee , G. (2021) Deep convolutional neural network based secure wireless voice communication for underground mines. Journal of Ambient Intelligence and Humanized Computing.

Full text not available from this repository. (Request a copy)

Abstract

A secure wireless voice communication system for underground miners is an essential gadget for efficient and safe mining. Voice over internet protocol is a proven solution for wireless communication in underground mines where other cellular and satellite networks cannot be deployed. However, the wireless network's security is the major issue for the reliable operation of the system. A secure voice communication system has been developed by integrating voice over internet protocol system and deep convolutional neural network (DCNN) based trained model. Experimental results indicated that voice recognition accuracy of the DCNN based developed model was 93.7% for the noiseless environment. In contrast, it was 82.1 and 79% for the existing K-nearest-neighbour (KNN) and support vector machine (SVM) algorithms, respectively. Voice recognition response time of the DCNN, KNN, and SVM algorithms was 178, 220, and 228 ms, respectively. Thus, deployment of the developed secure and robust voice communication system would improve safety and productivity in underground mines

Item Type:	Article
Uncontrolled Keywords:	Convolutional auto-encoder Deep convolutional neural network Underground mine VoIP Wireless communication
Subjects:	Instrumentation
Divisions:	UNSPECIFIED
Depositing User:	Mr. B. R. Panduranga
Date Deposited:	04 Jan 2021 11:15
Last Modified:	04 Jan 2021 11:15
URI:	http://cimfr.csircentral.net/id/eprint/2297

Actions (login required)

View Item