![Page 1: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/1.jpg)
Department of Computer Science
Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani
DTN Routing as a Resource Allocation Problem
![Page 2: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/2.jpg)
2
What are DTNs?
Delay/Disruption Tolerant Networks • end-to-end path may never exist• routing must use pair-wise transfers staggered over time
iX Z
Y
i
i
![Page 3: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/3.jpg)
3
Why useful?
Infrastructure expensive or nonexistent • e.g., Daknet, Kiosknet, OLPC
Infrastructure cannot be deployed• e.g., underwater, forests, outer space(!)
Infrastructure limited in reach e.g., Dieselnet, Cartel, Drive-thru-internet, VanLan
DTNs high delay, low cost, useful bandwidth
![Page 4: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/4.jpg)
4
Why challenging?
Wired/Mesh/MANETs Known topology Low feedback delay
• Retries possible
Primary challenge: finding a path to the destination under extreme uncertainty
DTNs Uncertain topology Feedback delayed/nonexistent
![Page 5: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/5.jpg)
5
Existing routing mechanisms Incidental
DTN routing mechanisms• Estimating meeting probability• Packet replication• Coding• Waypoint stores
• Prior knowledge• …
Metrics desired in practice• Minimize average delay• Maximize packets
meeting their deadlines• …
Incidental Routing• Effect of mechanism on routing metric unclearGoal: Design Intentional DTN Routing Protocol, RAPID
![Page 6: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/6.jpg)
6
Roadmap
Background and Motivation RAPID
Replication to handle uncertainty Utility-driven resource allocation Distributed algorithm
Deployment and Evaluation
![Page 7: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/7.jpg)
7
Replication to handle uncertainty
Replication can address• Topology uncertainty• High delay feedback
How to replicate when bandwidth is limited?
Naïve replication strategy: Flooding Risks degrade performance when resources limited
iX Z
Y
i
i
W
i
![Page 8: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/8.jpg)
8
Problem• Which packets to replicate given limited bandwidth to
optimize a specified metric
RAPID: Resource Allocation Protocol For Intentional DTN Routing
Routing as a resource allocation problem
![Page 9: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/9.jpg)
9
RAPID: utility-driven approach
RAPID Protocol (X,Y):
1. Control channel: Exchange metadata
2. Direct Delivery: Deliver packets destined to each other
3. Replication: Replicate in decreasing order of marginal utility
4. Termination: Until all packets replicated or nodes out of range
YX
Change in utility
Packet size
![Page 10: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/10.jpg)
10
Utility U(i): expected contribution of packet i to routing metric
Example 1: Minimize average delay• U(i) = negative expected delay of i
Example 2: Maximize packets delivered within deadline• U(i) = probability of delivering i within deadline
Example 3: Minimize maximum delay• U(i) = negative expected delay of i if i has highest delay;
0 otherwise
Translating metrics to utilities
![Page 11: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/11.jpg)
11
U(i) = -(T + D)• T = time since created, D = expected remaining time to deliver
Simple scenario• uniform exponential meeting with mean ¸• global view
Utility computation example
D = ¸
iX
D = ¸/2
iY
D =¸/3
iW
Z
![Page 12: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/12.jpg)
12
Utility computation example
Replicate i
Metric: Min average delay
Deadline of i < TDeadline of j = T1 > T
Metric: Max packets delivered within deadline
Replicate j
iX
j
Y j
W Z
![Page 13: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/13.jpg)
14
Roadmap
Background and Motivation RAPID
Replication to handle uncertainty Translating metrics to utilities Distributed algorithm
Deployment and Evaluation
![Page 14: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/14.jpg)
15
Distributed algorithm challenges
b
c
d
W
a
b
X
a
b
Y
Z
2sec
1sec5sec
Meeting times unknown
![Page 15: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/15.jpg)
16
Distributed algorithm challenges
Distributed control channel to build local view of unknowns
b
c
d
W
a
b
X
a
b
Y
Z
1pkt/2sec
3pkt/sec
2pkt/5sec
Meeting times unknown
Transfer size unknown
Replica locations unknown
(delivery unknown)
![Page 16: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/16.jpg)
17
Distributed control channel
b
c
d
W
a
b
X
a
b
Y
Z
1pkt/2sec
3pkt/sec
2pkt/5sec
Expected inter-meeting timeExpected transfer size
Known replica locations Expected “local” delay
per node
per packet
Expected delay of packet b
~ min(DW,b, DX,b, DY,b)
5 4 1
DX,b ~ 4sec
![Page 17: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/17.jpg)
18
RAPID recap
RAPID Protocol (X,Y):
1. Control channel: Exchange metadata
2. Direct Delivery: Deliver packets destined to each other
3. Replication: Replicate in decreasing order of marginal utility
4. Termination: Until all packets replicated or nodes out of range
![Page 18: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/18.jpg)
19
Is RAPID optimal ?
RAPID: No knowledge
Complete knowledge• NP Hard• Approximability lower bound √n
Partial knowledge• Average delay: arbitrarily far from optimal• Delivery rate: Ω(n)-competitive
Empirically, RAPID is within 10% of optimal for low load
DTN unknowns: Meeting schedule Packet workload Global view
![Page 19: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/19.jpg)
20
Roadmap
Background and Motivation RAPID
Replication to handle uncertainty Translating metrics to utilities Distributed algorithm
Deployment and Evaluation
![Page 20: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/20.jpg)
21
Deployment on DieselNet
![Page 21: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/21.jpg)
22
Results from deployment
Synthetic workload Deployed from Feb 6, 2007 until May, 14, 2007
![Page 22: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/22.jpg)
23
Results from deployment
Per day stats
Avg number of buses on road 19
Avg number of meetings 147.5
Bytes transferred (MB) 261.4
Average packet delay (min) 91.7
% packets delivered 88%
% meta data exchanged 1.7%
![Page 23: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/23.jpg)
24
Validating the simulator
Trace-driven simulator Simulation results within 1% of deployment
![Page 24: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/24.jpg)
25
Results: Mobility from DieselNet traces
![Page 25: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/25.jpg)
26
Results: Known mobility model
![Page 26: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/26.jpg)
27
Conclusions Intentional DTN routing feasible despite high uncertainty
• tunable to optimize a specific routing metric Simple utility-driven heuristic algorithm performs well in practice
• DTN routing problem fundamentally hard
Ongoing work• Application development on DTNs• Graceful degradation across mesh networks and DTNs
traces.cs.umass.edu
![Page 27: Department of Computer Science Aruna Balasubramanian, Brian Neil Levine, Arun Venkataramani DTN Routing as a Resource Allocation Problem](https://reader035.vdocuments.net/reader035/viewer/2022070411/56649f2f5503460f94c49d16/html5/thumbnails/27.jpg)
28
Questions?