Vehicle Ad-hoc Network (VANET) is a direct application of Mobile Ad-hoc Network (MANET). Nodes in VANET are vehicles that communicate using vehicle to vehicle (V2V) or vehicle to infrastructure (V2I). These types of communications have led to the emergence of various applications that provide safer driving. Due to the high changing of topology and frequent fragmentation of VANET, routing pack-ets in this type of network is a hard task. In this work, the authors deal with the well-known MANET proactive Optimized Link State Rout-ing protocol (OLSR). The deployment of OLSR in VANET gives the moderate performance; this is due to its necessity of constant ex-changing of control packets. The performance of OLSR is highly dependent on its parameters, thus finding optimal parameters configura-tions that best fit VANETs environment and improves the network is essential before its deployment. Therefore, this research proposes a modified Harmony Search optimization (HSO) by incorporating selection methods in its memory consideration; roulette wheel selection to obtain fine-tuned OLSR for high density and velocity scenario. The experimental analysis showed that the OLSR with the proposed ap-proach acquired promising results regarding packet delivery ratio, end-to-end delay and overhead when compared with previous approaches.