GVDSR: A DYNAMIC ROUTING STATEGY FOR VEHICULAR AD-HOC NETWORKS Michael Barros, Anderson Costa Federal Institute of Education, Science and Technology of Paraíba (IFPB) [email protected], [email protected]

Reinaldo Gomes Systems and Computing Department, Federal University of Campina Grande [email protected]

ABSTRACT

Routing in VANETs (Vehicular Ad-hoc Networks) is a developing area due the specific features of these networks. One of the challenges is to provide information even that the channel is rapidly modified by the network’s mobility. Mobility is a notable issue, since it has been faced by the logical part in the network. The network layer needs to ensure the stability of routing without relying on the mechanisms from the physical layer, and routing solutions have to consider different features for communication between different hosts/vehicles considering the constantly changing topology. This paper is presents a new routing strategy for VANETs called GVDSR (Generic Vehicular Dynamic Source Routing). This strategy is based on the Routing Architecture for VANET Communications presented in this paper. Experiments were performed in different scenarios to evaluate the proposed strategy and the obtained results were compared with AODV and DSDV protocols. These experiments showed the usefulness of the proposed strategy and the results indicate that our proposal gives markedly better performance on scenarios considered.

KEYWORDS Routing, VANETs, Ad-Hoc, Protocol, Architecture.

1. INTRODUCTION The specific characteristics of the network’s nodes and their communication patterns of VANETs will affect the routing strategy. The challenge of determine routes for information transport in vehicular networks is a complex work due the high nodes mobility in the network and instability of wireless links. The routing protocols for the communication between nodes are classified as: topological, geographical, opportunists and dissemination of information [Alves 2009]. The protocols based in topology found the best path between any other pair source-destination of the network. Typically, the best path is that offers the lowest cost according to the utilized metrics. These protocols can be proactive, reactive or hybrid. Position based routing (or geographical) is capable to provide more scalability in high mobility environments. In this approach, it’s not necessary to keep information about the routes of each node in the network [Gouqing 2008]. This type of routing assumes that the present elements in the network have some location system like GPS, as Galileo [Hein 2002]. The community research is realized in order to provide a greater diversity of features to these protocols. The developed protocols have the following features: carry large scales in the network on situations with high vehicles density to improve routing [Wang 2007]; support to the intense vehicle mobility, adapts quickly to the new topologies and enables a greater connection without a possible link breakage [Taleb 2007]; adaptation system to the constantly position exchange of nodes in the network [Ali 2009]; passivity [Xue

2008] and dynamic [Xi 2008]. Packets sent by proactive protocol have to update their paths database for the packets routing, in other words, the routing protocol in VANETs has to forward packets reactively. Based on this study, a Routing Architecture for Vehicular Communications was proposed in [Barros 2010]. This architecture contains relevant features which a routing protocol for VANET scenarios has to face. The Generic Vehicular Dynamic Source Routing is a new routing strategy for VANETs that follows this architecture. The GVDSR was implemented in the Network Simulator 2 [NS2 2010], and it was compared to others protocols for MANET with the objective of functional validation. The results show the contributions and the advantages for the GVDSR protocol and for the Routing Architecture. This paper is organized as follows: Section 2 introduces the Routing Architecture; the GVDSR protocol is presented in Section 3; the simulations are explained in Section 4 and the results in Section 5. The paper is concluded in Section 6.

Fig. 1. Routing Architecture for Vehicular Communications

Fig. 2. Used protocol on the architecture.

2. ROUTING ARCHITETURE FOR VEHICULAR COMMUNICATIONS Many routing protocols for VANETs were studied and based on that were observed that these protocols focus in specifics features and presents limited solutions. Some of the most significant features necessary for VANETs routing protocols:  Be dynamic from the origin. The protocol will be reactive, determining routes on-demand [Xi 2008].  Presents a good behavior in small scale situation, like in the streets where they have lower vehicles traffic; and in large scale, where the vehicles traffic is bigger like found in the avenues [Wang 2007].  Recognizes the networks topology even with the constantly change of nodes position, and do the routing without a lot of discarded packets [Ali 2009].  Provides greater time connectivity between vehicles, for the packets routing can be done. And also presents mechanisms for link breakage situations between vehicles [Taleb 2007]. To integrate these features in one routing strategy, an architecture was created to enable a new trend of routing protocols in VANETs. This new trend has to encompass a majority of VANETs features in a modularized way. The architecture is shown in the Figure 1. The Routing Architectures for VANETS presents the most significant features of the vehicular communication.

3. GENERIC VEHICULAR DYNAMIC SOURCE ROUTING - GVDSR Instantiating the Routing Architecture, a new protocol was developed: The Generic Vehicular Dynamic Source Routing (GVDSR). The GVDSR protocol presents all steps from the routing architecture applying VANETs existing protocols. The selected protocols that compose the GVDSR are shown in Figure 2. These protocols can be replaced in every moment, if there is a new protocol with a new technology. The GVDSR protocol is an extension of the DSR (Dynamic Source Routing) protocol [Johnson 2004], but it is designed for Vehicular Communications.

4. EXPERIMENTS The GVDSR was implemented in the Network Simulator 2 (NS2) [NS2 2010]. This simulator was selected for the facilities of implementing a new routing protocol, the wide documentation and tutorials founded in the Internet and because the simulations in the network layer is near from the real world. The main objective for implementing the GVDSR is the functional validation of the protocol and the architecture. Primary scenarios were chosen for the same reason. The scenarios are shown in Figure 6. There are four different scenarios, varying the number of the nodes and their direction. The nodes S and D are the source and the destination respectively. The I node is the intermediary node which is fixed, like roadside units. The S and D nodes move in 36km/h. The scenario c) is the unique on large scale for the distance between the final nodes, the other scenarios are also important for the validation of the routing strategy. The GVDSR protocol was compared with other protocols: AODV (Ad-Hoc On-Demand Distance Vector) [Perkins 2003] and DSDV (Destination Sequenced Distance Vector) [Perkins 1994]. Table 1. Experiments Parameters Factor Transceivers Propagation Model Antennas

Value IEE 802.11 standard Two Ray Ground Ominidirectional

Table 2. Logic Configuration Factor Traffic Protocol Application Protocol Connection Time

Value TCP FTP The all simulation time

The metrics for comparison were: Generated Packets, Dropped Packets and Total Throughput. For VANET scenarios, the nodes needs a connection with a low number of packets discarded and a large throughput for application, the metrics were chosen by these premises.

Fig. 6. Used Scenarios

5. RESULTS The results in the scenario a), as shown in Fig.7, present the GVDSR protocol adding some extra packets for routing. This strategy enables a higher discovery of information. The process of information discovery is appropriate for the evaluation of the specific features of VANETs. The number of dropped packets and the Total Throughput shows the effectiveness of the information discovery strategy. In the scenario b), Fig 8, the number of dropped packets for GVDSR is lowers than the others protocols. The Generated Packets for GVDSR is greater, but it is for the information discovery, that enables the decrease of the dropped packets. The throughput for applications will increase showing that the packet delivery ratio increased too.

Fig. 7. Results for scenario a)

Fig. 8. Results for scenario b)

Fig. 9. Results for scenario b)

Fig. 10. Results for scenario d)

The large scale scenario c), Fig. 9, is the one that shows the advantages of this architecture and the GVDSR. For large scale situation the GVDSR protocol acts greater than the others, improving the packets delivery ration with the decrease of packets number. This is ideal for VANETs scenario, with the increase of the distance, the networks needs to route information with the less number of packets discarded. For Fig 10, with the scenario d), the performance of the GVDSR protocol stays greater than the others protocols. The number of dropped packets is lower for GVDSR, enabling an increase of the throughput for applications since the increase of the vehicles speed.

6. CONCLUSION A new protocol for VANETs was presented in this paper: the GVDSR. This protocol was obtained as an instance of the Routing Architecture for VANET Communications. The protocol is subdivided in four steps. The steps are integrated in a modularized way for the selected algorithms being upgrade separately. The algorithm provides a dynamic source routing with checking the strength of the path link. The GVDSR provides different treatments for small or large scale situations. To decrease the link breakage the algorithm calculates a metric called LET, this metric shows the reliability of the link between two nodes. The GVDSR protocol was implemented in the NS2 simulator and compared with two protocols: AODV and DSDV. The results show that the GVDSR adds some packets for the strategy of information discovery. The number of dropped packets is the lowest in four different scenarios. For VANETs, the GVDSR protocol presents a greater packet delivery ratio and a greater throughput for the nodes. Other metrics need to be analyzed and evaluations for multimedia traffic are the future work.

REFERENCES Alves et al., 2009. Redes Veiculares: Princípios, Aplicações e Desafios. Proceedings of Simpósio Brasileiro de Redes de Computadores (SBRC) Minicourse. Gouqing, Z.; Dejun, M.; Zhong, X.; Weili, Y.; Xiaoyan, C., 2008. A survey on the routing schemes of urban Vehicular Ad Hoc Networks. Proceedings of 27th Chinese Control Conference. pp. 338 - 343. Hein, G.W., Godet, J., Issler, J.L., Martin, J.C.,Erhard, P., Lucas, R. e Pratt, T., 2002. Status of galileo frequency and signal design. Proceedings of International Technical Meeting of the Satellite Division of the Institute of Navigation ION GPS. pp. 266-277. Wang, W.; Xie, F.; Chatterjee, M., 2007. TOPO: Routing in Large Scale Vehicular Networks. Proceedings of IEEE 66th Vehicular Technology Conference. pp. 2106-2110. Taleb, T.; Sakhaee, E.; Jamalipour, A.; Hashimoto,K.; Kato, N.; Nemoto, Y., 2007. A Stable Routing Protocol to Support ITS Services in VANET Networks. In IEEE Transactions on Vehicular Technology. Volume 56, Issue 6, Part 1, pp. 3337-3347. Ali, S.; Bilal, S.M., 2009. An Intelligent Routing protocol for VANETs in city environments. Proceedings of 2nd International Conference Computer, Control and Communication. pp. 1-5. Xue, G.; Feng, J.; Li, M., 2008. A Passive Geographical Routing Protocol in VANET. Proceedings of IEEE Asia-Pacific Services Computing Conference. pp. 680-685. Xi , Sun; Li, Xia-Miao, 2008. Study of the Feasibility of VANET and its Routing Protocols. Proceedings of 4th International Conference on Wireless Communications, Networking and Mobile Computing. pp. 1 - 4. Barros, M.T.A.O.; Costa,A.F.B.F.; GOMES,R.C.M., 2010. Routing in VANETs: propose of a generic algorithm. Proceedings of Wireless Systems International Meeting (WSIM). Campina Grande, Brazil. Ns2 (The Network Simulator), 2010. http://www.isi.edu/nsnam/ns/ Johnson D.,Maltz D. e Yin-Chun H., 2004. The dynamic source routing protocol for mobile ad hoc networks (http://www.ietf.org/internet-drafts/draft-ietf-manet-dsr-10.txt). IETF Internet Draft. Mo Z., Zhu H.,Makki K., Pissinou N., 2006. MURU: A multi-hop protoccol for urban ad-hoc networks. Proceedings Third International Conference on Mobile and Ubiquitos System: Networking and Services. pp. 169-176. Perkins, C., Belding-Royer, E. e Das, S., 2003. {Ad hoc on-demand distance vector (AODV) routing (http://www.ietf.org/rfc/rfc3561.txt). IETF Internet Draft. Perkins, Charle E. and Pravin Bhagwat, 1994. Highly Dynamic Destination-Sequenced Distance-Vector Routing (DSDV) for Mobile Computers.