Lecture 6: Vehicular Computing and Networking Cristian Borcea Department of Computer Science NJIT
GPS & navigation system On-Board Diagnostic (OBD) systems DVD player Satellite communication 2
Internet Cellular Cellular Roadside infrastructure Roadside infrastructure Vehicle-to-vehicle Applications Accident alerts/prevention Real-time re-routing Entertainment Communication Cellular network (3G/4G) Vehicle to roadside (WiFi) Vehicle to vehicle (WiFi) 3
High node mobility Constrained nodes movements Obstacles-heavy deployment fields, especially in cities Large network size Can applications based on multi-hop communications work in such environment? 4
Introduction VANET applications: EZCab & TrafficView RBVT routing in VANET Real-time re-routing in vehicular networks 5
The nightmare! Phase1: Fight with other people for a cab Phase2: Call a dispatching center and wait, and wait 6
Need a cab Use mobile ad hoc networks of cabs to book a free cab Each cab has short-range wireless interface and GPS Prototype over Smart Messages 7
Discovery Phase B D Discover Free cab Busy cab A P D =0.5 P E =0.75 E P H =0.5 H P A =0.187 P B =0.375 P C =0.250 C F Busy cab P F =0.0 P G =0.5 G Free cab 8
Booking Phase B D H A P D =0.5 P E =0.75 =0.25 E P H =0.5 P A =0.187 =0.125 P B =0.375 =0.250 P C =0.250 C F Busy cab P F =0.0 P G =0.5 G Free cab 9
Updating phase B D H A P D A =0.5 =0.375 P E D =0.25 =0.5 P E =0.25 E P H =0.5 P A =0.125 =0.375 P B =0.25 P C =0.25 =0.75 C F Busy cab P F =0.0 P G =0.5 G P C =0.75 Free cab 10 P C =0. 50
Avgerage number of hops between a booked cab and its corresponding client Avgerage number of hops between a booked cab and its corresponding client Proactive books the closest cabs Average distance increases as the number of free cabs decreases 5 4 3 Flooding On-demand Proactive 10 9 8 7 6 Flooding On-demand Proactive 2 5 4 1 3 2 0 355 305 255 205 155 Number of free cabs from the 410 cabs 1 0 10 20 50 Number of cab requests per second 11
On foggy days What s in front of that bus? On rainy days What s behind the bend? 12
Provides dynamic, real-time view of the traffic ahead Initial prototype Laptop/PDA running Linux WiFi & Omni-directional antennas GPS & Tiger/Line-based digital maps Road identification software Second generation prototype adds Touch screen display 3G cards Possibility to connect to the OBD system 13
Problem: How to disseminate information about cars in dynamic ad-hoc networks of vehicles? Solution: broadcast all data in one packet (simple data propagation model) Use aggregation to put as much data as possible in one packet Aggregate data for vehicles that are close to each other Perform more aggregation as distance increases Maintain acceptable accuracy loss 14
Parameters Aggregation ratio: inverse of the number of records that would be aggregated in one record Portion value: amount of the remaining space in the broadcast message 1. Calculate region boundaries 2. Calculate merge thresholds 3. For every region, merge every two consecutive records closer than merge threshold Current Vehicle 15
High-density highway scenario Ratio-based aggregation performs best overall Visibility Accuracy 16
Introduction VANET applications: EZCab & TrafficView RBVT routing in VANET Real-time re-routing in vehicular networks 17
Examples of node-centric MANET routing protocols AODV, DSR, OLSR Frequent broken paths due to high mobility Path break does not always correspond to connectivity loss Performance highly dependent on relative speeds of nodes on a path S N1 D a) At time t S N2 N1 D b) At time t+δt 18
Examples of MANET geographical routing protocols GPSR, GOAFR Advantage over node-centric Less overhead, high scalability Subject to (virtual) dead-end problem S N1 N2 Dead end road D 19
Use road layouts to compute paths based on road intersections Select only those road segments with network connectivity Use geographical routing to forward data on road segments Advantages S I1 A B C I6 Source Path in header: I 8 -I 5 -I 4 -I 7 -I 6 -I 1 E I2 I4 I7 I3 I5 I8 D Greater path stability Lesser sensitivity to vehicles movements car Ij Intersection j Destination 20
RBVT-R: reactive path creation Up-to-date routing paths between communicating pairs Path creation cost amortized for large data transfers Suitable for relatively few concurrent transfers RBVT-P: proactive path creation Distribute topology information to all nodes No upfront cost for given communication pair Suitable for multiple concurrent transfers 21
Source broadcasts route discovery (RD) packet RD packet is rebroadcast using improved flooding Nodes wait before rebroadcasting packet a period inverse proportional to distance from sender If overhear another packet transmission, no need to rebroadcast Traversed intersections stored in RD header S Source I1 A B C N 1 Re-broadcast from B Re-broadcast from N 1 I2 I4 I3 I5 I6 E I7 I8 D car Ij Intersection j Destination 22
Destination unicasts route reply (RR) packet back to source Route stored in RR header RR follows route stored in RD packet S Source I1 A I2 I3 B C I4 I5 Path in reply packet header I 1 I 6 I6 E I7 I8 D I 7 I 4 I 5 Ij car Intersection j Destination I 8 23
Data packet follows path in header Geographical forwarding is used between intersections Path in data header I 1 S Source I 6 I 7 I 4 I1 A I2 I3 I 5 I 8 B C I4 I5 I6 E I7 I8 D car Ij Intersection j Destination 24
Dynamically update routing path Add/remove road intersections to follow end points When path breaks Route error packet sent to source Source pauses transmissions New RD generated after a couple of retries S Source I1 A B C N 1 Re-broadcast from B Re-broadcast from N 1 I2 I4 I3 I5 I6 E I7 I8 D Ij car Intersection j Destination 25
Unicast connectivity packets (CP) record connectivity graph Node independent topology leads to reduced overhead Lesser flooding than in MANET proactive protocols Network traversal using modified depth first search Intersections gradually added to traversal stack Status of intersections stored in CP Reachable/unreachable n-1 n CP generator I1 A B 2 I2 1 7 8 I3 C 3 6 I4 I5 5 9 I6 E 4 I7 I8 car i Step i Ij Intersection j 26
CP content disseminated in network at end of traversal Each node Updates local connectivity view Computes shortest path to other road segments I v1 I1 I2 I I3 v2 I v4 I 1 : I 2, I 6, I v1 I v3 I 6 : I 1, I 7 I 7 : I 6, I 4 I4 I5 I 4 : I 7, I 5, I v3 I 5 : I 4, I 8, I v4 I 2 : I 1, I v2 RU content I6 I7 I8 Reachability Ij Intersection j 27
RBVT-P performs loose source routing Path stored in every data packet header Intermediate node may update path in data packet header with newer information In case of broken path, revert to greedy geographical routing 28
hello packets used to advertise node positions in geographical forwarding hello packets need to be generated frequently in VANET High mobility leads to stalled neighbor node positions Presence of obstacles leads to incorrect neighbor presence assumptions Problems in high density VANET Increased overhead Decreased delivery ratio 29
Slight modification of IEEE 802.11 RTS/CTS Backward compatible RTS specifies sender and final target positions Waiting time is computed by each receiving node using prioritization function Next-hop with shortest waiting time sends CTS first Transmission resumes as in standard IEEE 802.11 (0.201ms) n 1 r RTS (0.0995ms) n s n (NULL) 2 (0.115ms) n 4 n 3 CTS n 5 n 6 (a) RTS Broadcast and Waiting Time Computation n s n 4 r n s n 1 n2 r n 5 n 3 n 6 n 1 Data n 2 (b) CTS Broadcast n 5 n 4 n 3 n 6 n s n 4 ACK n 1 n2 (c) Data Frame r n 5 n 3 n 6 D D D D 30
Function takes 3 parameters Distance from sender to next-hop (d SNi ) Distance from next-hop to destination (d i ) Received power level at next-hop (p i ) Weight parameters α 1,2,3 set a-priori Their values determine weight of corresponding parameter 31
RBVT-R with source selection using hello packets vs. self-election Distributed next-hop self-election Increases delivery ratio Decreases end-to-end delay 32
Average delivery ratio (%) Average delivery ratio (%) 150 nodes 250 nodes 100 100 90 90 80 80 70 60 50 40 30 AODV GPSR RBVT-P OLSR GSR RBVT-R 70 60 50 40 30 AODV GPSR RBVT-P OLSR GSR RBVT-R 20 20 10 10 0 0.5 1 1.499 2 3.003 4 4.505 5 Packet sending rate (Pkt/s) 0 0.5 1 1.499 2 3.003 4 4.505 5 Packet sending rate (Pkt/s) RBVT-R has the best delivery ratio performance RBVT-P improves in medium/dense networks The denser the network, the better the performance for road-based protocols 33
End-to-end delay (Seconds) End-to-end delay (Seconds) 150 nodes 250 nodes 5 5 4.5 4.5 4 4 3.5 AODV 3.5 AODV 3 2.5 2 GPSR RBVT-P OLSR GSR 3 2.5 2 GPSR RBVT-P OLSR GSR 1.5 RBVT-R 1.5 RBVT-R 1 1 0.5 0.5 0 0.5 1 1.499 2 3.003 4 4.505 5 0 0.5 1 1.499 2 3.003 4 4.505 5 Packet sending rate (Pkt/s) Packet sending rate (Pkt/s) RBVT-P performs best Consistently below 1sec in these simulations RBVT-R delay decreases as the density increases Fewer broken paths 34
Why? How long is the current route going to last? Does it make sense to start a route discovery? Can a 100Mb file be successfully transferred using the current route? Is it possible to estimate the duration of a path disconnection? How to estimate path characteristics (connectivity duration/probability)? Simulations are specific to geographical area Analytical models based on validated traffic models are preferred 35
DTMC-CA derives probabilistic measures based only on vehicle density for a traffic mobility model Microscopic Cellular Automaton (CA) freeway traffic model DTMC-MFT generalizes the approach used by DTMC-CA to any vehicular mobility model Focuses on macroscopic information of vehicles rather than their microscopic characteristics Values predicted by models are similar to simulation results from validated CA traffic model 36
Enhancing route maintenance of RBVT-R How long should the source wait when a route breaks Network overhead decreases up to 50% Delivery ratio and latency remain similar 37
Introduction VANET applications: EZCab & TrafficView RBVT routing in VANET Real-time re-routing in vehicular networks 38
Use global real-time traffic knowledge to dynamically guide drivers to alternative routes Goals: lower travel time for each driver, avoid congestion Byproducts: reduce fuel consumption, pollution Use smart phones for instantly deployable solution 39
Re-routing triggered when congestion predicted on certain road segments Congestion predicted using Segment-specific short term historical data (speed, volume) Static information (i.e., road capacity and speed limit) Speed-volume equations Select of vehicles to be re-routed according to utility function E.g., remaining travel time Selected vehicles provided with alternative paths that lower current predicted travel time Paths don t have to be the shortest Goal: avoid moving congestion from one segment to another 40
Privacy Reduce frequency with which drivers report their position, cloak destination Robustness System works with low penetration rate & in presence of drivers who ignore guidance Accurate real-time traffic view traffic Adapt number and frequency of reports submitted by smart phones to balance accurate global traffic view with privacy Effective real-time guidance Push guidance to drivers fast to allow them enough time to switch on new route Scalability Low communication overhead 41
MANET: best privacy protection and quickly predict congestion in small regions Localized, non-optimal decisions Peer-to-peer: same privacy benefits as MANET and acquire a global view of the traffic Difficult to provide fast guidance; significant overhead Off-load some computation to vehicles: server distributes global traffic view to vehicles, which make local decisions Better privacy & scalability Server + MANET: vehicles make collaborative decisions 42
EZCab 1. http://cs.njit.edu/~borcea/papers/percom05.pdf TrafficView 2. http://dl.acm.org/citation.cfm?id=1031487 3. http://cs.njit.edu/~borcea/papers/vtcsp04.pdf RBVT routing 4. http://cs.njit.edu/~borcea/papers/ieee-tvt08.pdf 5. http://cs.njit.edu/~borcea/papers/acm-tomacs10.pdf 43
1. Two decades of mobile computing 2. Infrastructure support for mobility 3. Mobile social computing 4. People-centric sensing 5. Programming mobile ad hoc networks 6. Vehicular computing and networking 7. Privacy and security in mobile computing Location privacy Location authentication Trusted ad hoc networks 44