The experimental evaluation of wireless and mobile networks is a challenge that rarely substitutes simulation in research works. This statement is even more evident in vehicular communications, due to the equipment and effort needed to obtain significant and realistic results. In this paper, key issues in vehicular experimental evaluation are analyzed by an evaluation tool called AnaVANET, especially designed for assessing the performance of vehicular networks. This software processes the output of well-known testing tools such as ping or iperf, together with navigation information, to generate geo-aware performance figures of merit both in numeric and graphical forms. Its main analysis capabilities are used to validate the good performance in terms of delay, packet delivery ratio and throughput of NEMO, when using a road-side segment based on IPv6 GeoNetworking.

1. AnaVANET:
an experiment and visualization tool
for vehicular networks
TRIDENTCOM 2014
Guangzhou, People's Republic of China
6th, May, 2014
Manabu Tsukada Inria,fr => UTokyo,jp
José Santa UMU,es
Satoshi Matsuura NAIST, jp
Thierry Ernst Mines ParisTech, fr
Kazutoshi Fujikawa NAIST, jp

2. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Outline
2
Background
Issue and requirements for VANET evaluation
Evaluation methodology
AnaVANET
Evaluation of NEMO over IPv6 GeoNetworking

3. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Vehicular Networking
3
Intelligent Transportation Systems (ITS)
Systems to achieve road safety, traffic efficiency and comfort
Cooperative ITS
Systems where multiple ITS Stations share information to
achieve better road safety, traffic efficiency and comfort
Vehicle, Roadside, Central and Personal ITS Stations
Communication
Vehicle ITS Station
Central ITS Station
Personal ITS Station
Roadside ITS Station
✓Road safety
✓Traffic efficiency
✓Comfort

4. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Network Layer Protocols for vehicular Networking
4
Vehicle to Vehicle (V2V)
wireless communications in dynamic topologies
without any infrastructure
VANET or MANET
Vehicle to Infrastructure (V2I)
global connectivity of nodes to the Internet
Internet Mobility (Mobile IPv6, NEMO)
AODV OLSR GeoNetworking,
C2CNet
Mobile IPv6 NEMO
Network Layer Protocols for vehicular communications
Topology based
Position based
Reactive Proactive
Network mobilityHost mobility
Infrastructure basedInfrastructure less (VANET or MANET)

5. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Issues on vehicular networks evaluation
5
Common end-to-end evaluation tools such as
ping6 and iperf
1. Unawareness of communication path
2. Unawareness of per-hop network
performance in the communication path
3. Unawareness of vehicles movement
Sender（ping6, iperf）
Receiver

6. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Requirements on vehicular network evaluation
6
1. Path detection
2. Communication performance in links
3. Geographical awareness
4. Intuitive visualization
5. Independence from network protocols
6. Independence from devices
7. Adaptation to various scenarios
8. Easiness for data collection

7. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Overview of Evaluation Methodology
7
The evaluation goals are to analyze which
testing conditions affect which
performance indicators using
target network protocols.
Distance Static Urban Highway
Packet size,
Send rate
TCP window size,
Max segment size
Packet size,
send interval
Testing Scenarios
Data Flows
Hardware (CPU, Memory), Antenna, Wireless setting(frequency, data rate)Testbed platform
Performance Indicators
AnaVANET
PDR, throughput, Jitter,
Hop count
Throughput
RTT, PDR, Hop
count
Infrastructure-basedInfrastructure less
(NEMO)(OLSR / C2CNet / IPv6 GeoNetworking)
Testing
Conditions
Target Network Protocols
(Examples)
Number of Vehicles
UDP TCP ICMPv6

8. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Testbed platform
8
V2V and V2I network
ARs and MRs
OS: Linux 2.6.29.6
Wireless configuration
•frequency 2.422Ghz
•Data rate 6 Mbits/s
NEMO is supported
GeoNetworking is
supported
Wireless
Wireless
MR1
MNN1
cable
MR2
MNN2
cable
MR3 MR4
Vehicular Network
Infrastructure Network
AR1
AR2
IPv6 Internet
HA1

9. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Testing Scenarios
9
Location Scenarios
INRIA Paris-Rocquencourt installations
Number of vehicles
Up to four vehicles are considered in our case
Mobility Scenarios Distance Static
Urban Highway
parked ~10km/h
Obstacle Obstacle
parked parked
parked parked
packet
~100km/h
~100km/h
~100km/h
Obstacle
~30km/h
~30km/h
~30km/h

10. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Data Flows and Performance Indicators
10
Data Flows
UDP
TCP
ICMPv6
Performance Indicators
Packet Delivery Ratio
Round Trip Time
Bandwidth
Jitter
Geographic metric
Hop count
Speed
Position of cars
Distance between cars
Sender（ping6, iperf）
Receiver

11. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Overview of the AnaVANET system
11
Generate packets with UDP, TCP and ICMPv6
Capture packets in all the nodes
Trace them packet by packet
MNN1
(Sender)
MNN2
(Receiver)MR1 MR4 MR2
Cable CableWireless Wireless
UDP / ICMPv6
traffic generation
iperf / ping6
log
tcpdump
log
GPS
log
tcpdump
log
GPS
log
tcpdump
log
GPS
log
iperf
log
AnaVANET
(optional)
(optional)
XML
statistics
Packet
trace
GnuplotWeb font-end
(Google maps)AnaVANET Web viewer
Graphshttp://anavanet.net/
MR3
tcpdump
log
GPS
log
Wireless

12. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Output examples
12
Google maps shows the
performance indicators with
Movement
Distance
Obstacles
!
!
!
!
!
!
!
!
Graphs generated by gnuplot
shows as follow
0 50 100 150 200 250 300
NEMOstatus
Time (seconds)
Fail
Succes
0
20
40
60
80
100
PDR(%)
Echo Request via AR2
0
20
40
60
80
100
PDR(%)
Echo Request via AR1
0
100
200
300
400
0
5
10
15
20
25
30
Distance(m)
RTT(ms)
Distance between MR and AR1
Distance between MR and AR2
RTT

13. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
IPv6 GeoNetworking using NEMO (ICMPv6)
13
64 bytes in 0.5s interval
RTT is about 5ms
Binding registrations are
lost in the south road
Handover
without packet loss
120m from AR2, 50m from AR1
0 50 100 150 200 250 300
NEMOstatus
Time (seconds)
Fail
Success
0
20
40
60
80
100
PDR(%)
Echo Request via AR2
0
20
40
60
80
100
PDR(%)
Echo Request via AR1
0
100
200
300
400
0
5
10
15
20
25
30
Distance(m)
RTT(ms)
Distance between MR and AR1
Distance between MR and AR2
RTT
AR2
AR1
MR
0
20
40
60
80
100
RTT (ms)
RTT
Packet loss
Binding registration Success
Binding registration Fail
RTT (ms)
http://anavanet.net/demo-vienna/?analysis=1296754401

14. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
IPv6 GeoNetworking using NEMO (UDP)
14
1250 Bytes packet with 1
Mbps rate
UDP packets are lost during
4 seconds in handover
Disconnection time after the
binding registration failure
was seven seconds (= 1 +
2 + 4)
0 20 40 60 80 100 120 140 160 180
NEMOstatus
Time (seconds)
Fail
Success
0
20
40
60
80
100
PDR(%)
PDR to AR2
0
20
40
60
80
100
PDR(%)
PDR to AR1
0
100
200
300
400
0
100
200
300
400
500
Distance(m)
Throughput(Kbits/sec)
Distance between MR and AR1
Distance between MR and AR2
Throughput
AR2
AR1
MR
0
20
40
60
80
100
PDR (%)
PDR to AR1
PDR to AR2
Binding registration Success
Binding registration Fail
PDR (%)
http://anavanet.net/demo-vienna/?analysis=1296759090

15. www.anavanet.net AnaVANET: an experiment and visualization tool for vehicular networks
Conclusions and future works
15
Conclusions
Analysis of evaluation of vehicular networks
Evaluation methodology
•Analysis of metric impacts the network performance
•Performance indicators
Design and Implementation of AnaVANET
Evaluation of IPv6 GeoNetworking using NEMO
Future work
Link layer extension i.e. channel quality, load ratio
Support for multicast data flows
Evaluation of real Cooperative ITS application

16. Thank you for attention
Manabu Tsukada
<tsukada@hongo.wide.ad.jp>