Heuristic Optimization for Reliable Data Congestion Analytics in Crowdsourced eHealth Networks

Abstract

Reliable data congestion analytics in crowdsourced eHealth networks becomes particularly important, especially in big data era, because of wide adaption of ubiquitous crowdsourced healthcare participants. Since a crowdsourced eHealth network has intermittent connectivity to its remote healthcare provider, researchers usually use some well-studied networks to model the novel network, but data congestion analytics is still a big problem in most intermittent connecting networks. In most cases, data congestion analytics may be realized by fixing the number of forwarded copies, but sometimes, it cannot suit the changing network environments well. This problem could be solved by modifying packet forwarding conditions dynamically through detecting real-time network environment. Based on this idea, in this paper, an optimized routing algorithm named RSW (reduced variable neighborhood search-based spray and wait) is proposed. In the algorithm, nodes will exchange and store each other's buffer status during their communication, based on which, current network environments will be evaluated and quantified as a real-time threshold. Then, spray and wait adapts the threshold for data congestion control. Simulation shows that the proposed algorithm increases data packet delivery probability, and optimize the overhead ratio dramatically, which can be up to ten times lower than that of standard algorithm.

Keywords

• Data analytics
• congestion control
• optimization
• crowdsourced eHealth networks