SOURCE:

Faculty: Engineering
Department: Electronic And Computer Engineering

CONTRIBUTORS:

Nwokolo, C. P.
Inyiama, H. C.

ABSTRACT:

The rate of crime has greatly been on the increase and the security of life, property and infrastructure has been an issue of national concern over the years and increasingly so all over the world. This research focused on developing an On-demand Cloud-Based Remote Security Video Sensing System (OD-RGRSVS) which is activated by only phone call or Short Message Service (SMS). IP surveillance cameras with infrared features have been used as both active and passive remote sensing devices for efficient video capture and storage. The system comprises a control station which receives surveillance service request signal termed save our souls (sos) from subscribers (who are mobile clients with telephone handsets) and a client station (in a fixed geographical location) equipped with surveillance cameras where the video capture takes place. The control station stores location data of clients and has a personal computer (PC) interface, with a graphical user interface (GUI) which was developed using Visual Basic.NET to visually manage the interface. The system has been designed to render Remote Video Sensing as a Service (RVSaaS) with cloud-based storage. The video processing event was characterized as a Poisson process modeled as a Markov chain. The cloud-based model was developed and simulated using Riverbed Modeler 17.5 and the network performance analysis was carried out on the metrics of network latency, throughput and resource utilization for assurance of Quality of Service (QoS). The primary backup took about 2ms while the secondary took about 10ms delay, and a network throughput of 100packets/sec was generally obtained without virtualization which significantly increased to 900packets/sec (primary) and 1100packets/sec (secondary) over time with virtualization. A prototype of the surveillance system comprising microcontroller units and global system for mobile communications (GSM) modules has also been built. Validation of results was carried out using CloudAnalyst on Java-based CloudSim network simulator. The system was able to drastically overcome video storage limitations by incorporating on-demand feature, with GSM-only activation, and virtualization at the cloud storage end as compared to Digital Video Recorder (DVR) and other conventional online backups that fail to leverage on storage virtualization.