Renu Dalal And Manju Khari. Persuasive Simulation of Optimized Protocol for OppNet. Dynamic Systems and Applications 30 (2021) No.5, 865-899

https://doi.org/10.46719/dsa202130512

ABSTRACT.
The new paradigm approach to provide communication in heterogeneous network with distinct communication technologies is known as Opportunistic Network (OppNets). This network permits the mobile nodes to cooperate in diversified social communication like micro-blogs and content dissemination. Due to discontinuous connections, lower bandwidth, and arbitrary nature of OppNets, establishing a new protocol for OppNets routing is a primary challenge. This paper introduces peculiar protocol for OppNets, it integrates the terrific features of random forest decision tree algorithm and firefly nature inspired algorithm. This protocol is trained by few features of Direct delivery protocol, after this machine learning algorithm Random Forest (RF) is used to find the reliable relay nodes (more than one) for transmitting the packet from source to destination node. After RF phase, only one best appropriate node is selected based on firefly optimization algorithm for transferring the packet. This protocol performs better than other protocol like first contact, epidemic, and direct delivery; in terms of message delivery probability, average hop count, overhead ratio, and average latency.

Keywords –Machine Learning, Opportunistic Network, Routing Protocol, Random Forest, Firefly Algorithm, ONE Tool.