Packet routing is important for vehicular delay tolerant networks (VDTNs). Opportunistic routing algorithms based on historical records are insufficiently accurate in forwarder selection due to movement randomness of vehicles. Trajectory-based routing algorithms tackle vehicle movement randomness but cannot be directly used in VDTNs due to the dependence on APs. In this paper, we develop a distributed trajectory-based routing algorithm (called MobiT) for VDTNs. This non-trivial task faces three challenges. First, vehicle trajectories must be sufficiently collected. Second, the trajectories cannot be updated frequently due to limited resources of the repository nodes. Third achieving high routing performance even with partially collected trajectories. Our real trace study lays the foundation of the design of MobiT. Taking advantage of different roles of vehicles, MobiT uses service vehicles that move in wide areas to collect vehicle trajectories, and rely on the service vehicles and roadside units (called schedulers) for routing scheduling. By using regular temporal congestion state of road segments, MobiT schedules the packet to arrive at a roadside unit prior to the destination vehicle to improve routing performance. Furthermore, MobiT leverages vehicles’ long-term mobility patterns to assist routing. Our trace-driven simulation and real experiments show the effectiveness and efficiency of MobiT.
• Data forwarding and routing in mobile opportunistic networks have gained much attention recently.
• The packet is forwarded at the direction maximizing the utility. SADV lets packets wait at intersections until the path with minimum delay is available
• Tie et al. proposed the Robust Replication Routing (R3), which unifies mesh, MANET, DTN routing paradigms by predicting the distribution of link delays. Kong et al. proposed a frequency divided instantaneous neighbour estimation system for vehicular networks.
• Schwartz et al. presented guidelines for the design of data dissemination in vehicular networks. These works all rely on historical information to predict future encounter. However, as indicated in the chosen forwarder may not meet the destination vehicle in large-scale vehicular networks due to the movement randomness of some vehicles, which impairs routing accuracy and efficiency.
We consider a VDTN with n vehicles denoted by Ni(i = 1, 2, . . ., n) and make following assumptions.
(1) Each vehicle has a Dedicated Short Range Communication (DSRC) device . When two vehicles are within each other’s communication range, an encounter happens.
(2) Each vehicle is equipped with a navigation system, which generates trajectory consisting of future positions and estimated arrival times, and road maps.
(3) The area in the VDTN is partitioned into multiple sub district with equal number of landmarks. Following, we assign a center landmark for each sub-district. Each intersection is installed with an RSU which uses DSRC for communication.
• Success rate
• Average delay
• Average number of information queries
• Average vehicle memory usage
TRACE DATA ANALYSIS ALGORITHM:
Trajectory of a service vehicle & Determine the relay vehicle & Utilize the friend vehicle
HARDWARE AND SOFTWARE REQUIREMENT:
Operating System : Linux
Simulation Tool : NS2
CPU type : Intel Pentium 4
Clock speed : 3.0 GHz
Ram size : 512 MB
Hard disk capacity : 80 GB
Monitor type : 15 Inch color monitor
Keyboard type : Internet keyboard
CD -drive type : 52xmax
J. Broch, D. A. Maltz, D. B. Johnson, Y.-C. Hu, and J. Jetcheva, ``A performance comparison of multi-hop wireless ad hoc network routing protocols, in Proc. MobiCom, 1998, pp. 85_97.