With the advancement of wireless communication technology among mobile devices, mobile ad hoc networks (MANETs) are becoming more common. This results in the growing need for efficient communication protocols and routing algorithms in MANETs. One important obstacle in achieving this is the ability to predict the time varying network states of a mobile ad hoc network (MANET). In this thesis, we model a MANET using dynamic graphs and study ways of predicting delay features. We propose different approaches, 1) two event-based approaches and 2) a time-based approach, to conceptualize the problem. We use long-short term memory (LSTM) neural networks to predict the one-hop delay values of any two devices in a given mobile network.
"},{"label":"Degree","value":["Bachelor of Arts with Departmental Honors"]},{"label":"Author-supplied Keywords","value":["Mobile ad hoc network","Dynamic graph","Long-short term memory neural network"]},{"label":"Administrative Note","value":"Drupal Node ID: 23996