Most disruption-tolerant networking protocols focus on mere contact and intercontact characteristics to make forwarding decisions. We propose to relax such a simplistic approach and include multi-hop opportunities by annexing a node’s vicinity to its network vision. We investigate how the vicinity of a node evolves through time and whether such knowledge is useful when routing data. By analyzing a modified version of the pure WAIT forwarding strategy, we observe a clear tradeoff between routing performance and cost for monitoring the neighborhood. By observing a vicinity-aware WAIT strategy, we emphasize how the pure WAIT misses interesting end-to-end transmission opportunities through nearby nodes. For the datasets we consider, our analyses also suggest that limiting a node’s neighborhood view to four hops is enough to improve forwarding efficiency while keeping control overhead low. www.phe-neau.com

1. The Strength of Vicinity
Annexation in Opportunistic
Networking
Tiphaine Phe-Neau*, Marcelo Dias de Amorim*,
and Vania Conan+
* UPMC Sorbonne Universités, France
+ Thales Communications & Security, France
The Fifth IEEE International Workshop on Network Science for
Communication Networks (NetSciCom 2013), Torino, Italy
April, 19th 2013

2. PEOPLE NETWORK
=

3. Disruption-Tolerant
Networks (DTN)
t = t1 t = t2 t = t3
3

4. Now
t0 t1
F
F G
B G
B
A A D
C D C
E E
4

5. K-vicinity
K in N*
A
1-vicinity (k = 1)
5

6. K-vicinity
K in N*
A
2-vicinity (k = 2)
6

7. Now with The Vicinity
t0 t1
F
F G
B G
B
A A D
C D C
E E
No Infinite
Waiting Delay Waiting Delay
7

8. Infocom05 (41) Rollernet (61)
RandomTrip
(20)
Unimi (48) Community (50) 8

9. Which K-Vicinity?
Infocom05
Neighbor distribution
k=1
k=2
0,7 1,5 3 k=3 1,4
k=4
k=5
1,5
2,3 k=6
k=7
3
4,4 k=8
No meaningful evolution
beyond k = 4. 9

10. Average waiting Times
Average WAITing times (seconds)
18300
contact 18232s
500 κ=2
κ=3
κ=4 40% 18200
κ=5
400
κ=6
18100
300 80% 1%
18000
200
17900
100 57% 40% 17792s
17800
0
Random Trip Community Infocom05 Rollernet Unimi*
Higher K brings
better waiting delays. 10

11. Message Overhead: Regular
Infocom05
100000
10000
No (messages)
κ=3
1000 κ=4
κ=2
100
κ = 5+
Contact
10
1
0 5000 10000 15000 20000 25000 30000 35000
Time t (seconds)
Overhead increase with K until k = 4.
11

12. Message Overhead: On
Demand
Infocom05
κ=2
κ=3
1000
N0 (messages)
100
κ=5+
contact
10 κ=4
1
10000 15000 20000 25000 30000 35000
Time t (seconds)
Sensing the 3+-vicinity is cheaper
than the 2-vicinity. 12

13. Take THIS Away !
• Sensing direct contacts in DTN is
not enough.
• k-vicinity reduces waiting delays.
• Network knowledge is costly so
limit it to k=4 !
13

14. www.phe-neau.com