Relay node selection using future movement predication for opportunistic network

• To develop a mobility-aware routing protocol that uses Artificial Neural Networks (ANNs) to predict future node movements in opportunistic network environments. • To design a relay node selection mechanism based on routing table insights, enabling efficient and reliable data forwarding despite intermittent connectivity. • To evaluate the performance of the proposed protocol against existing routing schemes using key metrics such as packet delivery ratio, end-to-end delay, and communication overhead. Opportunistic Networks (OppNets) function effectively in environments where stable connectivity and traditional networking infrastructure are lacking or inconsistent. Due to the highly dynamic nature of these networks, accurately predicting node movement and selecting appropriate relay nodes are essential for improving routing efficiency and resource management. This work introduces a protocol named Relay Node Selection using Future Movement Prediction (RNSFMP), which combines mobility forecasting through Artificial Neural Networks (ANNs) with a Connecting Table-based Relay Node Selection (CT-RNS) mechanism. The mobility prediction model employs ANNs to analyze past movement patterns and estimate future node positions. Based on this foresight, the CT-RNS algorithm uses locally maintained routing data to identify optimal relay candidates for forwarding data toward the destination. To evaluate its performance, the proposed RNSFMP protocol was tested against established routing schemesMaxProp, Epidemic, and PRoPHET. Metrics such as packet delivery ratio, end-to-end delay, and transmission overhead were used in the comparison, and the results indicate that RNSFMP offers improved data delivery efficiency in intermittently connected environments.