DTN

Page

DTNs more often than not require communications between mobile nodes. To correctly identify patterns in contacts and predicting them with accuracy is a research challenge. I devised an algorithm that predicts contacts to an average accuracy of 95%  on the tested number of nodes. Predictions were based on the history of contacts and this was reduced to an equation using numerical approximation. The program then learnt dynamically and predicted node contacts with steady accuracy for higher node contact instances.