The Impact of Rogue Nodes on the Dependability of Opportunistic Networks

Opportunistic Networks (OppNets) are an extension to the classical Mobile Ad hoc Networks (MANETs) where the network is not dependent on any infrastructure (e.g. Access Points or centralized administrative nodes). OppNets can be more flexible than MANETs because an end to end path does not exist and much longer delays can be expected. Whereas a Rogue Access Point is typically immobile in the legacy infrastructure based networks and can have considerable impact on the overall connectivity, the research question in this project evaluates how the pattern and mobility of a rogue nodes impact the dependability and overall "Average Latency" in an Opportunistic Network Environment. We have simulated a subset of the mathematical modeling performed in a previous publication in this regard. Ad hoc networks are very challenging to model due to their mobility and intricate routing schemes. We strategically started our research by exploring the evolution of Opportunistic networks, and then implemented the rogue behavior by utilizing The ONE (Opportunistic Network Environment, by Nokia Research Centre) simulator to carry out our research over rogue behavior. The ONE simulator is an open source simulator developed in Java, simulating the layer 3 of the OSI model. The Rogue behavior is implemented in the simulator to observe the effect of rogue nodes. Finally we extracted the desired dataset to measure the latency by carefully simulating the intended behavior, keeping rest of the parameters (e.g. Node Movement Models, Signal Range and Strength, Point of Interest (POI) etc) unchanged. Our results are encouraging, and coincide with the average latency deterioration patterns as modeled by the previous researchers, with a few exceptions. The practical implementation of plug-in in ONE simulator has shown that only a very high degree of rogue nodes impact the latency, making OppNets more resilient and less vulnerable to malicious attacks.

Level dc:type.qualificationlevel

masters

Grantor dc:publisher.institution

National University of Ireland Maynooth

Year dc:date.issued

2015

Author dc:creator

• Agha, Faizan

Language dc:language

en