Performance testing of OLSR using mobility predictions

OLSR (Optimized Link State Routing) protocol routing is a proactive routing protocol presented by the IETF MANET (Mobile Ad-hoc NETwork) working group for ad-hoc networks. The common drawback of all table-driven protocols in MANETs is the large routing overhead for the creation of routing tables. In order to tackle this issue, OLSR uses the Multipoint Relaying protocol (MPR), which main use is to limit the flooding of OLSR routing packets. In a previous study, we illustrated how MPR could benefit from mobility predictions. We introduced the Kinetic Multipoint Relaying (KMPR) protocol which selects kinetic multipoint relays based on the overall nodes' predicted degree in the absence of trajectory changes. In this paper, we propose to study the application of the KMPR protocol for OLSR. We show that thanks to KMPR's improved topology knowledge, the route error ratio (RER) and packet delivery ratio (PDR) are consequently improved. OLSR is also able to benefit from the low fraction of messages required my KMPR in order to maintain its topology. The emphasis of this paper is to show that thanks to the use of mobility prediction, OLSR's performances may be significantly improved, at virtually no extra cost in term of routing overhead. Experiments demonstrate a cut by 6 of the RER, while an increase of the PDR of 60\% compared to the regular OLSR.