Fragmented Data Routing Based on Exponentially Distributed Contacts - - PowerPoint PPT Presentation

UCLA ENGINEERING Computer Science

Fragmented Data Routing Based on Exponentially Distributed Contacts in Delay Tolerant Networks

Tuan Le, Qi Zhao, Mario Gerla Computer Science, UCLA {tuanle, qi.zhao, gerla}@cs.ucla.edu

UCLA ENGINEERING Computer Science

Outline

• Background
• Motivation
• Proposals
• Protocol Design
• Evaluation
• Conclusion

Slide 2 / 16 2/20/19

UCLA ENGINEERING Computer Science

Background

Delay-Tolerant Mobile Ad-Hoc Networks

• Sparsely connected
• End-to-end paths are rarely available due to node mobility

2/20/19 Slide 3 / 16

UCLA ENGINEERING Computer Science

Motivation

• Pedestrians with hand-held devices communicate via

Bluetooth

• High-speed vehicles communicate via WiFi (802.11g)
• Both cases have short contact duration: several seconds

2/20/19 Slide 4 / 16

UCLA ENGINEERING Computer Science

Motivation

• Existing works assume unfragmented data
• Large data not fit short contact duration
• Abort entire transmission
• Useless retransmission of entire data

2/20/19 Slide 5 / 16

20 MB 5 sec contact duration

UCLA ENGINEERING Computer Science

Proposals

• Single-copy data fragmentation
• Break data into small chunks
• Transmitted at various contacts
• Compute direct/indirect delivery probability to

successfully deliver all chunks via multiple contacts

2/20/19 Slide 6 / 16

C1 C2 C3 Fragments

UCLA ENGINEERING Computer Science

One-Hop Delivery Probability

• Probability message i is successfully delivered from s to d

during the nth meeting is the joint probability of 3 events

• Message i does not expire before the nth meeting
• d does not receive all parts of message i during the first n-1 meetings
• d receives the remaining parts of message i during the nth meeting

Slide 7 / 16 2/20/19

UCLA ENGINEERING Computer Science

One-Hop Delivery Probability

2/20/19 Slide 8 / 16

• Joint probability Pi(n)
• One-hop (direct) delivery probability
• and – kth inter-contact time and contact duration time between s and d
• Ri – remaining lifetime of message i
• = x B – amount of data sent from s to d during kth contact
• B – communication bandwidth between two nodes
• Wi – size of message i

Xs,d

k

Y s,d

k

Pi =

∞

X

n=1

Pi(n)

Zs,d

k

Y s,d

k

Pi(n) = P ⇣ 0 <

n

X

k=1

Xs,d

k

< Ri ⌘ · P ⇣ 0 ≤

n−1

X

k=1

Zs,d

k

< Wi ⌘ · P ⇣

n

X

k=1

Zs,d

k

≥ Wi ⌘

UCLA ENGINEERING Computer Science

Two-Hop Delivery Probability

• Probability message i is successfully delivered from s to v

during their nth meeting and from v to d during their mth meeting

• Two-hop (indirect) delivery probability

2/20/19 Slide 9 / 16

Pi(n, m) = P ⇣ 0 <

n

X

a=1

Xs,v

a

+

m

X

b=1

Xv,d

b

< Ri ⌘ ·P ⇣ n−1 X

a=1

Zs,v

a

< Wi ⌘ · P ⇣

n

X

a=1

Zs,v

a

≥ Wi ⌘ ·P ⇣ m−1 X

b=1

Zv,d

b

< Wi ⌘ · P ⇣ m X

b=1

Zv,d

b

≥ Wi ⌘

Pi =

∞

X

n=1 ∞

X

m=1

Pi(n, m)

UCLA ENGINEERING Computer Science

Routing Strategy

2/20/19 Slide 10 / 16

• S computes one-hop delivery probability Ps to d and two-hop

delivery probability Pvi to d via each neighbor vi

• Pv = max(Pv1, Pv2, …, Pvm)
• If Pv > Ps and no chunk of message i resides at another node, s

forwards all parts of message i to v that fit contact duration

s v d Pv Ps Current contact Direct delivery probability Pv > Ps Indirect delivery probability

UCLA ENGINEERING Computer Science

Evaluation

• ONE simulator 1.5.1
• Cabspotting trace: 536 taxis collected over 30 days in

San Francisco Bay Area

• Each node has 5 source messages of same size

intended for random destinations

• Messages have a homogenous TTL value
• Nodes have infinite buffer capacity

2/20/19 Slide 11 / 16

UCLA ENGINEERING Computer Science

Evaluation

Evaluate the following schemes:

• Epidemic routing: flood messages (multi-copy, unfragmented)
• PROPHET: select relay with higher delivery probability to the

destination (single-copy, unfragmented)

• BubbleRap: bubble up messages to node with high global ranking.

Once reach community, bubble down using local ranking (single-copy, unfragmented)

• MEED: select relay with lower minimum expected delay (single-copy,

unfragmented)

• Fragmented data routing (FDR) : select relay with max(one-hop, two-

hop delivery probability) (single-copy, fragmented)

2/20/19 Slide 12 / 16

UCLA ENGINEERING Computer Science

Performance comparison 10KB

2/20/19 Slide 13 / 16

0.5 1 2 4 6 8 10 12 TTL (days) 0.2 0.4 0.6 0.8 1 Delivery ratio

Epidemic FDR PROPHET MEED BubbleRap

0.5 1 2 4 6 8 10 12 TTL (days) 0.5 1 1.5 2 2.5 Average delay (days)

Epidemic FDR PROPHET MEED BubbleRap

Delivery ratio Average delay

Performance comparison with messages of 10 KB

UCLA ENGINEERING Computer Science

Performance comparison 20MB

2/20/19 Slide 14 / 16

0.5 1 2 4 6 8 10 12 TTL (days) 0.2 0.4 0.6 0.8 Delivery ratio

Epidemic FDR PROPHET MEED BubbleRap

0.5 1 2 4 6 8 10 12 TTL (days) 0.5 1 1.5 2 2.5 3 3.5 Average delay (days)

Epidemic FDR PROPHET MEED BubbleRap

Delivery ratio Average delay

Performance comparison with messages of 20 MB

UCLA ENGINEERING Computer Science

FDR Performance

2/20/19 Slide 15 / 16

20 0.2 0.4 15 15

Delivery ratio

0.6

Message size (MB)

0.8 10 10

TTL (days)

1 5 5

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

20 0.5 1 15 15

Average delay (days)

1.5

Message size (MB)

2 10 10

TTL (days)

2.5 5 5

0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Delivery ratio Average delay

Performance of FDR with varied message sizes from 10KB to 20MB

UCLA ENGINEERING Computer Science

Conclusion

• Forwarding decision using message fragmentation is

aware of contact duration and message size

• Consider both direct and indirect delivery probability
• FDR improves delivery rate and delay by up to 37%

and 35%, respectively

Slide 16 / 16 2/20/19

UCLA ENGINEERING Computer Science

Thanks!

Q & A