Scalable Models for Redundant Data Flow Analysis in Online Social Networks

Sharing and reception of identical video contents amongst the users of various online social media applications induce enormous amount of redundant data traffic flow across Internet layer, various planes of Service provider’s network and also increase in the data consumption of the end users. The primary aim of this research is to reduce the routing of redundant data across social media by means of a novel procedure based on the validation of pre-existence of contents at the receiver’s end prior to the actual video content transmission. A proactive middleware is proposed to handle the enormous growth of social media traffic transparently and hence a distributed Web based framework has been implemented for real-time traffic flow analysis. As part of the Service Oriented Architectural model, various services are defined to fetch a frame from each initiated video content, to validate the pre-existence at receiver’s end and to take appropriate decision for acceptance or deferral of the actual video transfer. The middleware which has been developed for redundant data flow traffic analysis is further extended to cloud services to address the scalability issues. The statistical analysis carried out by the proposed models depicted 27% of reduction in the total initiated traffic with adequate amount of data saving for the end users. The overhead associated in the transmission of the first frame of the actual video is also analyzed and it is observed that overhead is led by the substantial amount of gains achieved by the proposed model.