Research Area:  Fog Computing

Abstract:

Sensor cloud is an integral component for smart computing infrastructure. Cloud servers are largely used to store and process sensor data. For mission critical applications use of only wireless sensor network results in provisioning of service in a small area and the use of a long distant remote cloud servers increase delay that degrades the Quality of Service. Further, geospatial information differs over regions. Thus storing and processing the data of all regions inside the cloud data centres may not be efficient with respect to response time (latency), energy consumption etc., which are crucial factors for mission critical applications. To overcome these limitations, we propose multi-sensor geo-fog paradigm. We consider defense sector in our work as mission critical application. For energy optimized services with minimal delay fog computing has been used, where the intermediate devices process the data. The proposed paradigm will offer fast and energy-efficient processing of defense related sensor and geospatial data. A mathematical model of the paradigm is developed. The sensor and geospatial data processing and analysis take place inside the fog device. If abnormality is detected in the data or emergency situation occurs, then shortest path to the victim region is determined using intelligent K* heuristic search algorithm. The simulation results demonstrate that the proposed fog based network scenario reduces energy consumption, average jitter and average delay by 12–15%, 10–14% and 9–11% respectively than the cloud based network. The simulation results demonstrate that saving about 20% of resources increases the performance for priority user whereas the resource availability for the normal users is not compromised.

Keywords:  
Energy-aware
multi-sensor
geo-fog
mission critical applications

Author(s) Name:  Moumita Mishra, Sayan Kumar Roy, Anwesha Mukherjee, Debashis De, Soumya K. Ghosh & Rajkumar Buyya

Publisher name:  Springer

DOI:  10.1007/s12652-019-01481-1

Volume Information:  volume 11, pages; 3155–3173 (2020)

Paper Link:   https://link.springer.com/article/10.1007/s12652-019-01481-1