Formal Modeling on Improvements of Route Discovery Algorithm for AODV Using UPPAAL

Mobile ad-hoc network (MANET) is a technology that enables a set of wireless mobile devices or routers to connect to each other, by establishing a temporary network, with no network infrastructure or any fixed management available. MANET's network topology constantly changes, therefore it is a challenge to preserve the routes between nodes. Hence, routing protocol existed as an instrument to meet these challenges. Ad-Hoc On-Demand Distance Vector (AODV) is a part of the on-demand routing protocol for MANET as nodes select a route on one condition when there is a node that wants to send data packets. One of the disadvantages of an on-demand routing protocol is a high probability for a node to lose communication link to the other since the unpredictability of changes on the network topology occurs continuously. This paper presents a methodical approach for verifying the route discovery algorithm of AODV variants using model checking with UPPAAL model checker, R-AODV, and OCR-AODV. In UPPAAL, the simulation tool can execute an exhaustive search for every state in a network of automata. In our analysis, we considered a broken link scenario during the route discovery process. Our UPPAAL message sequence chart shows that Reverse AODV achieved route discovery despite an intermediate node losing its transmission on the network while OCR-AODV requires rebroadcasting the RREQ message. These results may be used to suggest further additions to the route discovery algorithm in AODV.