New technologies for both internal and external vehicle communication are emerging for real time safety critical systems as well as entertainment, and wireless communication for remote user application in the local and wide area. By incorporating these technologies into the vehicle new and innovative functions can be provided to the passengers. The research aims to provide seamless integration of these different functional units in the in-vehicle network. Inter-vehicle communication is also being considered by using wireless technology (in particular Bluetooth) for high mobility Ad-hoc networking.

Integration, In-Vehicle Network

An ad hoc network consists of wireless mobile nodes forming a temporary network, without the need for a fixed infrastructure. Many studies in ad hoc networking mainly concentrate on the packet routing methodologies. Some of the most popular routing protocols are DSDV, TORA, DSR and AODV. But these protocols mainly perform well in scenarios where node mobility and topology changes are very slow. The following section consists of a brief description of each of these protocols.

The performance of these protocols improves as mobility of the nodes decreases. For example, Josh Broch [J.Broch .A performance comparison of multihop wireless ad hoc network routing protocols., 4th ACM/IEEE International Conf. On Mobile Computing and networking, pages, 85-97, Oct 1998] did a comparative study of these protocols, adding mobility to the nodes. The simulations were conducted using 10, 20, and 30 sources, and the maximum node mobility was restricted to 20 m/s (72 km/h).

All protocols deliver a greater percentage of the originated data packets, when there is little node mobility (i.e. large pause times), converging to 100% delivery when there is no node motion. DSR and AODV perform very well, delivering over 95% of the data packets, regardless of the mobility rate. But here the maximum speed is just 20 m/s. For Inter vehicle communication applications, the vehicle speeds will be much greater than this; also the number of participating nodes will be much higher. Further more the use of Bluetooth, imposes certain restrictions by which the network topologies are formed in a different manner. While these approaches are certainly feasible it may not provide the most efficient solution. For example due to the packet size limitations of the Bluetooth baseband layer, all MANET style solutions will require fragmentation and reassembly of packets at each relay node. This calls for new methods by which communication between neighbouring nodes and routing can be achieved.

In Bluetooth the basic network that is formed is called a piconet. Piconets can join together by sharing bridging nodes to form large networks to form scatternets. The problem of defining scatternet formation criteria is itself an open research issue that is heavily dependent on the envisioned application. Routing of packets in such a network will then depend on how the network topology or scatternets are formed. There have been a few studies in this area; the following section discusses some of these protocols.

All the study in Bluetooth ad hoc networks mainly deal with networks in which nodes are mainly static and the topology changes are very slow. The high mobility aspects of the nodes as in the case for Inter Vehicle communication are not considered. Furthermore the mobility patterns of the participating nodes affect the protocols in different ways. Hence, the protocols being developed need to be studied using vehicular traffic models.

Using the Bluetooth connection establishment times derived earlier (average connection establishment time), the communication time window between two vehicles can be calculated. Consider the figure below:

Connection Window

Here Vehicle A is travelling at 97 Km/h and Vehicle B is travelling at 113 km/h. The difference of 16 km/h equates to approximately 4.5 m/s. After coming into range, Vehicle B would need to travel 200m more than Vehicle A for it to exceed the Bluetooth range. Using these speed values the two vehicles would be in range for a period of approximately 44 seconds. During this time period vital information can be exchanged between the vehicles depending on the envisaged application.

Mobile Vehicle Links

Positional data is very important for the internetworking between vehicles and is exchanged with time stamps. This is an important parameter in any considered application due to the high mobility of all the vehicles. For example a vehicle may move in and out of networks, but using positional information and speed it had acquired at a certain time and comparing it with current time it would be able to predict the vehicles location in its immediate vicinity and try to establish direct links with the relevant ones.

Another method to deal with Ad hoc networking for inter-vehicle communication is by using the concept of .clusters.. In Bluetooth clusters are formed automatically when piconets are created. Here each vehicle or node belongs to a particular cluster having a unique ID, which would be the master.s address.

Ad-hoc Vehicle Clusters

Each piconet (cluster) can be thought of as having a virtual perimeter around it. The diagram above shows two types of perimeter, which could be constructed around the devices. Here the master of each piconet would transfer information (tables as in the previous case) amongst its slaves. The slaves falling on the edge of the perimeter would be made the bridging nodes (relay nodes), such that inter-piconet communication can take place. Each device would need to be aware of the neighbouring piconets, whose ID and perimeter values would be communicated through the bridging nodes. Knowledge of the neighbouring piconets would help in nodes migrating to other piconets to form quick connection since the masters ID would be known.

All the above issues are subject to investigation. The research uses a combination of simulating tools (BlueHoc for ns-2) and practical tests on Ericsson and CSR Bluetooth development kits.