Vehicle ad hoc network
A vehicle ad hoc network ( Vehicular Ad Hoc Network, VANet ) is a mobile ad hoc network (MANet) whose nodes are vehicles. As a rule, motor vehicles are regarded as nodes of a VANet . As with a MANet, it is a self-organizing and decentralized network. In contrast to MANets, vehicle ad hoc networks are only the subject of research in computer science . There is no established national or international vehicle ad hoc network for vehicle-to-vehicle communication .
Network architecture and characteristics
A vehicle ad hoc network consists of mobile nodes or vehicles (English. Car ) and base stations along the road or traffic network (English. Road-Side-Units, RSU ). In VANets, a connection is referred to as vehicle-to-vehicle communication (Car to Car, C2C) and vehicle-to-base station communication (Car to Infrastructure, C2I). A vehicle ad hoc network is therefore usually also viewed as a hybrid of purely infrastructure-based, cellular communication and ad hoc communication.
A VANet must meet special requirements that result from its area of application and that determine the network characteristics.
- The nodes of a VANet move at different speeds, but usually very high compared to typical MANets. This highly dynamic topology results in correspondingly short connection times (slow fading ). With an average vehicle speed of 100 kilometers per hour and a maximum transmission or reception range of, for example, 200 meters, oncoming vehicles can only have a direct connection for approx. 4 seconds. Concepts for VANet-based safety applications envisage vehicle speeds of up to 250 km / h.
- Changing environmental conditions lead to constantly changing signal propagation and rapid fading or complete shading . This leads to frequent connection interruptions.
- Nodes (vehicles) usually move according to corresponding patterns, e.g. B. the local conditions (e.g. road network, motorway, one-way street), customary movement patterns (e.g. acceleration, turning) and collective patterns (e.g. traffic jam). From this information for the movement prediction can be made. B. can be used in a corresponding VANet routing algorithm .
Routing
Just like a MANet , a VANet requires flexible routing algorithms that differ from the procedures established in wired or cell-based systems. The convergence of the algorithm and the handling of a highly dynamic network topology and asymmetrical connections play a special role in a VANet . The problem of power supply that is otherwise typical for routing in MANets does not play a major role in VANets. In addition to topology-based routing , geo-based routing protocols assume the regular availability of position information for all network nodes.
Topology-based routing
With topology-based routing in VANets, a distinction is made between flat routing and hierarchical routing . If the routing algorithm is based on a flat network structure, the network is viewed as a single region by each network node. This means that each node forwards data to all nodes within range (its neighbors that are ready to receive). The essential advantage of such flat routing protocols is that they do not require any specialized network nodes and corresponding organizational mechanisms. A lack of coordination leads to highly dynamic topologies such as However, in the case of VANets, for example, the maintenance of the routing table is too expensive . A complete storage of all participants in local routing tables, strictly following the idea of flat routing, cannot be efficiently implemented in a VANet due to the high number of participants . Therefore, flat routing protocols are often only considered in limited scenarios in VANet implementations and simulations. Hierarchical routing counteracts this problem by organizing network nodes into groups and dividing functions by means of selected network nodes inside and outside the group. The grouping takes place according to geographical proximity or according to reception quality in explicit clusters, which can also be organized hierarchically. Each cluster is represented by a cross-cluster communicating node (cluster head).
Topology-based routing protocols in VANets also fall into two categories:
There are also hybrid procedures consisting of proactive and reactive routing.
Many proactive routing protocols originate from the Link State (LS) protocol. The proactive routing protocols in MANets include B. Dynamic Destination-Sequenced Distance-Vector (DSDV), Wireless Routing Protocol (WRP), Global State Routing (GSR), Fisheye State Routing (FSR), Topology Broadcast Based on Reverse Path Forwarding (TBRPF) or Optimized Link State Routing ( OLSR). The costs of proactive routing in VANets are very high due to the constant change in the routing table entries, of which only a few are actually used. The frequent sending of packets with updated routing information results in a communication overhead that is too high for VANets with typically limited bandwidth. Of the proactive protocols, only the hierarchical FSR is generally considered in VANet implementations .
The reactive routing protocols include B. Ad-hoc On-demand Distance Vector (AODV), Dynamic Source Routing (DSR), Associativity Based Routing (ABR), Lightweight Mobile Routing (LMR) and the Temporally Ordered Routing Algorithm (TORA). AODV and DSR in particular can be found in many VANet implementations . Reactive routing protocols determine a route only when required (on demand), which reduces the communication overhead and administrative effort for maintaining the routing tables . These are significant advantages for use in VANets, but the reactive determination of the route selection results in a delay at each affected node. This can be compensated in MANets by a buffer for recently determined routes, but cannot be implemented as efficiently in VANets due to the highly dynamic network topology.
Geo-assisted routing
Due to the widespread use and availability of GPS receivers in vehicles, geo-supported routing protocols were focused on in the development of VANet concepts from the start.
Well-known VANet research projects
In the FleetNet research project of the German Federal Ministry of Education and Research (BMBF), approaches were developed between September 2000 and December 2003 in order to exchange data on multihop routes between nearby vehicles.
Network on Wheels (NOW) is a research project also supported by the BMBF, which was founded in 2004 by Daimler AG , BMW AG, Volkswagen AG , the Fraunhofer Institute for Open Communication Systems as well as NEC Deutschland GmbH and Siemens AG . The aim of this project is to solve key technical questions in the area of communication protocols and data security .
Secure Vehicular Communication (SEVECOM) is an EU-funded project that aims to investigate and guarantee security and anonymity in vehicle networks. The project includes Daimler AG , Robert Bosch GmbH , Ulm University and Trialog as project coordinator.
VANet implementations
- NEC Car2X Communication Stack is a Linux implementation of a commercial VANet stack. The NEC Car2X stack supports IPv4 , IPv6 and geographic routing.
- U2VAS , the U NIVERSITY U lm VA Net S tack (written: U 2 VAS ) is a Java implementation of a VANET stacks. U2VAS has a completely modular structure and currently implements position-based routing (multi-hop CGGC ), visualization, as well as authenticity and authentication using a public key infrastructure (PKI).
- The ACUp Communication Framework is a C ++ implementation of a high-performance VANet stack for research purposes, especially in the field of driver assistance and active safety . The modular, expandable Car2X framework enables high data rates and has a client API for Windows and Linux.
Web links
Individual evidence
- ↑ Hierarchical routing protocols for MANETs ( page no longer available , search in web archives ) Info: The link was automatically marked as defective. Please check the link according to the instructions and then remove this notice. - Chair of Computer Networks and Internet, University of Tübingen
- ↑ Ad Hoc Mobile Wireless Networking (PDF; 167 kB) - Heritage Institute of Technology, Kolkata
