vanetS Irem Nizamoglu Computer Science amp Engineering Outline Motivation Epidemic Protocols EpiDOL Parameter Optimization Performance Results amp Adaptivity Features Conclusion Outline ID: 141133
Download Presentation The PPT/PDF document "Epidemic density adaptive Data dissemina..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Epidemic density adaptive Data dissemination exploiting opposite lane in vanetS
Irem NizamogluComputer Science & EngineeringSlide2
Outline
MotivationEpidemic ProtocolsEpiDOL
Parameter Optimization
Performance Results &
Adaptivity
Features
ConclusionSlide3
Outline
MotivationEpidemic Protocols
EpiDOL
Parameter Optimization
Performance Results &
Adaptivity
Features
ConclusionSlide4
Motivation
Increase the safety of passengers,Disseminating emergency packets or road condition information efficiently,
Decreasing the fuel consumption and air pollution.
Longest recorded traffic jam in the world (260 km)-
Shangai
/China.Slide5
Outline
MotivationEpidemic Protocols
EpiDOL
Parameter Optimization
Performance Results &
Adaptivity
Features
ConclusionSlide6
Epidemic Protocols
Probabilistic information dissemination which does not require any knowledge of the network topologies.Suitable for VANETs;
Provides intelligence while reducing contentions and collisions.
Not require infrastructure support.
Fits well with the non-deterministic nature of VANETs (highly dynamic and unpredictable topology changes).Slide7
Epidemic Protocols
Protocol
Disconnected
Network Problem
Reality of the traces
Minimize Delay
Edge-Aware
[1]
-
✔
-
DV-CAST
[2]
✔
-
✔
DAZL[3]✔--EpiDOL✔✔✔
[1]
M
.
Nekovee
, “Epidemic algorithms for reliable and efficient information dissemination in
vehicular ad hoc networks,”
Intelligent Transport Systems, IET, vol. 3, no. 2, pp. 104 –110,
june
2009
.
[2]
O
.
Tonguz
, N.
Wisitpongphan
, and F.
Bai
, “
Dv
-cast: A distributed vehicular broadcast protocol for vehicular ad hoc networks,” Wireless Communications, IEEE, vol. 17, no. 2, pp. 47 –57,
april
2010.
[3]
R
.
Meireles
, P.
Steenkiste
, and J. Barros, “
Dazl
: Density-aware zone- based packet forwarding in vehicular networks,” in Vehicular Networking Conference (VNC), 2012 IEEE, pp. 234–241. Slide8
Outline
MotivationEpidemic Protocols
EpiDOL
Parameter Optimization
Performance Results &
Adaptivity
Features
ConclusionSlide9
EpiDOL
Goal: Maximize throughput while disseminating data in a certain area and keeping the overhead and delay below a certain level of threshold.Key properties:
Defining flags for packet dissemination direction and vehicles’ movement direction, deciding intelligent transmission,
Using opposite lane in an epidemic manner efficiently,
Decreasing collision rate by using density adaptive probability functions
p
same
,
p
opposite
and
p
sameToOpp
.
Including range adaptivity feature that utilizes channel busy ratio and reception rate.Slide10
EpiDOL
Performance Metrics:End-to-End Delay: Time taken for packet transmission from source to nodes in
the range of dissemination distance.
Throughput
:
R
ate
of successfully received packets by all nodes
within
dissemination distance.
Opposite
Lane:
H
ow
many times opposite lane nodes resend the packets that are taken from the original
side.
Overhead: The number of duplicate packets received during the simulation.Slide11
EpiDOL
df
: direction flag
of
: original flagSlide12
EpiDOLSlide13
Outline
MotivationEpidemic Protocols
EpiDOL
Parameter Optimization
Performance Results &
Adaptivity
Features
ConclusionSlide14
Parameter Optimization
For
density adaptive probability functions;
However
, as a result of the analysis best α value is different in the same and the opposite sides.Slide15
Parameter Optimization
For the same directional probability best αsame is chosen as 15 where;
max throughput>90% such that
eed
<0.06 s & overhead<
0.07.Slide16
Parameter Optimization
For the opposite directional probability best αopposite is chosen as 21 where;
max throughput>97% such that
eed
<0.08 s & overhead<0.1
.Slide17
Parameter Optimization
For calculation of PsameToOpposite,
we need to specify
backwardValue
.Slide18
Parameter Optimization
To achieve 90% throughput in lower densities. backwardValue
> 9.
Considering
overhead values for several different vehicle densities, the optimum
backwardValue
is determined as
11
. Slide19
Outline
MotivationEpidemic Protocols
EpiDOL
Parameter Optimization
Performance Results &
Adaptivity
Features
ConclusionSlide20
Performance Results & Adaptivity Features
Background Traffic:
1 KB sized FTP packets with 1, 0.1, 0.01 second frequency.Slide21
Performance Results & Adaptivity Features
Background Traffic (con’t
):Slide22
Performance Results & Adaptivity Features
Range
Adaptivity
:
Included
a transmission range
adaptivity
feature to achieve the maximum possible throughput at different densities and data rates.
Channel Busy Ratio (CBR
)
: ratio
of the busy time of the channel over all time
.
0.4 < CBR < 0.7
0.3 sec/packet
0.5 sec/packet
1 sec/packetSlide23
Performance Results & Adaptivity Features
Range
Adaptivity
(
con’t
):
Limits are specified from previous graphs.Slide24
Performance Results & Adaptivity Features
Range
Adaptivity
(
con’t
):
Reception
rate
:
successfully received packets in 1 second period of time.
1< Reception Rate < 1.5Slide25
Performance Results & Adaptivity Features
Range
Adaptivity
(
con’t
):
Between 1 and 1.5, we have high throughput.Slide26
Performance Results & Adaptivity Features
Range Adaptivity (
con’t
):Slide27
Performance Results & Adaptivity Features
Range Adaptivity (
con’t
):Slide28
Performance Results & Adaptivity Features
Comparative Results:
Compared
EpiDOL
and
EpiDOL+Adaptivity
with protocols in literature; DV-CAST, Edge-Aware and DAZL.Slide29
Performance Results & Adaptivity Features
Comparative Results (
con’t
):Slide30
Outline
MotivationEpidemic Protocols
EpiDOL
Parameter Optimization
Performance Results &
Adaptivity
Features
ConclusionSlide31
Conclusion
At low densities, achieved more than the %90
throughput.
EpiDOL
handled the disconnected network problem
.
At
high densities,
throughput achieved by
EpiDOL
is better than the others.
Indicates
that broadcast storm problem did not effect our protocol due to its probabilistic density adaptive functions. Slide32
Conclusion
Unless the background traffic is heavy, EpiDOL is not significantly affected .
The last version of the
adaptivity
function improves throughput %25 in high densities while comparing with raw
EpiDOL
.
Future work; consider more complicated highway structures.Slide33
Publication
I. Nizamoglu, S. C. Ergen and O.
Ozkasap
, "
EpiDOL
: Epidemic Density Adaptive Data Dissemination Exploiting Opposite Lane in VANETs
", EUNICE Workshop on Advances in Communication Networking, August 2013
. [
pdf
|
link
]
In preparation to submission (Journal):
Epidemic
Density
A
daptive Data Dissemination Exploiting Opposite Lane in VanetsSlide34
THANK YOU
Irem
Nizamoglu
:
inizamoglu@ku.edu.tr
Wireless Networks Laboratory:
http://wnl.ku.edu.tr