140 likes | 272 Views
Knowledge-Based Opportunistic Forwarding in Vehicular Wireless Ad Hoc Networks. Jason LeBrun, Chen-Nee Chuah, Dipak Ghosal, Michael Zhang University of California, Davis Presented by Long Vu. Scenario and Assumptions. Mobility of nodes can’t be controlled. GPS device usage.
E N D
Knowledge-Based Opportunistic Forwarding in Vehicular Wireless Ad Hoc Networks Jason LeBrun, Chen-Nee Chuah, Dipak Ghosal, Michael Zhang University of California, Davis Presented by Long Vu
Scenario and Assumptions Mobility of nodes can’t be controlled GPS device usage How to decide to (or not to) forward packets to neighbors if global topology is unknown or change rapidly?
Five Forwarding Schemes • NoTalk • Broadcast • Location-based • MoVe • MoVe-LookAhead
NoTalk M • Removes message M • One copy of M exists in the entire network Hello Response M
Broadcast M • Keeps message M • Many copies of M exists in the entire network Hello Response M M
Location-based • Prefer nodes closer to the destination to forward the packet
Motion Vector (MoVe) N1 C1
MoVe-LookAhead N1 N2 C1 C2
Simulation Settings • Piecewise-linear movement • Follow city street structure • Stand 10 (s) at waypoints
Results for Real Bus Trace • Fixed, known routes • Location-based routing algorithm outperforms other algorithms • No actual plot
Conclusion • Inconsistent results • Experimental traces: MoVE LookAhead performs best (exchanging velocity improves data dissemination) • Real bus traces: Location-aided performs best • Unknown routes of mobile nodes • How to ensure packet dissemination to destination? (although nodes exchange packet)