1. A Delay-Tolerant Network Architecture for Challenged Internets
Kevin Fall
September 16, 2018
Anshul Kantawala

2. Challenged Networks Terrestrial mobile networks
Unexpected partitions due to node mobility or RF interference
Periodic, predictable partitions
e.g. Commuter bus acting as store and forward switch
September 16, 2018
Anshul Kantawala

3. Challenged Networks (cont.)
Exotic Media Networks
Near-Earth satellites, very long-distance radio (deep space) etc.
High latencies with predictable interruption
Outage due to environmental conditions
Predictably available store and forward network service – e.g. low-earth orbiting satellites
September 16, 2018
Anshul Kantawala

4. Challenged Networks (cont.)
Military Ad-Hoc Networks
Operate in hostile environments
mobile nodes, environmental factors or intentional jamming cause disconnections
Data traffic may be pre-empted by higher priority voice traffic
Strong infrastructure protection requirements
September 16, 2018
Anshul Kantawala

5. Challenged Networks (cont.)
Sensor networks
Limited end-node power, memory and CPU capability
Thousands or millions of nodes per network
Communication scheduled to conserve power
Interfaced to other networks using proxy nodes
September 16, 2018
Anshul Kantawala

6. Current Solutions Link-repair approach Use proxy agents
Engineer problem links to appear similar to regular links
Use proxy agents
Attach challenged networks at edges using proxy agents
Does not provide a general way to use these networks for data transit
September 16, 2018
Anshul Kantawala

7. Characteristics of Challenged Networks
Path and Link characteristics
Network architectures
End System characteristics
September 16, 2018
Anshul Kantawala

8. Path and Link characteristics
High latency, low data rate
e.g. 10 kbps, 1-2 second latencies
Asymmetric data rates
e.g. remote instruments – large return channel, small uplink for device control
Protocols should be terse and dynamic control functions performed open-loop or hop-by-hop
September 16, 2018
Anshul Kantawala

9. Path and Link characteristics
Disconnection
Non-faulty disconnections
Motion
Predictable: satellite passes, bus acts as router
Random: motion of nodes/routers, interference
Low-duty-cycle operation
Routing subsystem should not treat predictable disconnections as faults and can use this information to pre-schedule messages
September 16, 2018
Anshul Kantawala

10. Path and Link characteristics
Long queueing times
Conventional networks rarely greater than a second
Challenged network could be hours or days due to disconnection
September 16, 2018
Anshul Kantawala

11. Network Architectures
Interoperability considerations
Networks may use application-specific framing formats, data packet size restrictions, limited node addressing and naming etc.
Security
End-to-end approach not attractive
Require end-to-end exchanges of keys
Undesirable to carry traffic to destination before authentication/access control check
September 16, 2018
Anshul Kantawala

12. End System Characteristics
Limited longevity
Round-trip time may exceed node’s lifetime making ACK-based policies useless
Low duty cycle operation
Disconnection affects routing protocols
Limited resources
Affects ability to store and retransmit data due to limited memory
September 16, 2018
Anshul Kantawala

13. Can we use TCP/IP? Transport layer (TCP) Network layer (IP) Routing
High latency and moderate to high loss rates severely limit TCP’s performance
Network layer (IP)
Performance affected by loss of fragments
Routing
High latency will cause current routing protocols to incorrectly label links as non-operational
September 16, 2018
Anshul Kantawala

14. Proxies and Protocol Boosters
Proxies and protocol boosters are inherently fragile
Increase system complexity if mobility is frequent
May require both directions to flow through the proxy – fail for asymmetric routing
Application proxies have limited re-use abilities and may fail to take advantage of special resources of the proxy node
September 16, 2018
Anshul Kantawala

15. Delay Tolerant Message-Oriented Overlay Architecture
September 16, 2018
Anshul Kantawala

16. Abstraction Message switching Overlay architecture
Use message aggregates or “bundles”
Allows network’s path selection and scheduling functions a-priori knowledge of the size and performance requirements of data transfers
Overlay architecture
DTN will operate over existing protocol stacks and provide a gateway when a node touches two or more dissimilar networks
September 16, 2018
Anshul Kantawala

17. Regions and DTN Gateways
DTN gateways are interconnection points between dissimilar network protocol and addressing families called regions
e.g. Internet-like, Ad-hoc, Mobile etc.
DTN gateways
Perform reliable message routing
Perform security checks
Store messages for reliable delivery
Resolve globally-significant name tuples to locally-resolvable names for internal destined traffic
September 16, 2018
Anshul Kantawala

18. Name Tuples Two variable length portions Region name Entity name
Globally-unique hierarchically structured region name
Used by DTN gateways for forwarding messages
Entity name
Resolvable within the specified region, need not be unique outside it
E.g. { internet.icann.int, }
September 16, 2018
Anshul Kantawala

19. Class of Service Similar to the Postal service
Delivery priority: low, ordinary, high
Notifications of mailing, delivery to receiver and route taken
Reliable delivery using custody transfer at each routing hop
September 16, 2018
Anshul Kantawala

20. Path Selection and Scheduling
End-to-end path routing path cannot be assumed to exist
Can solve a multicommodity flow optimization problem using approximate algorithms, since the protocol is message based
September 16, 2018
Anshul Kantawala

21. Custody Transfer Two types of message nodes Custody Transfer
Persistent (P) and non-persistant (NP)
P nodes assumed to contain persistent memory storage and participate in custody transfer
Custody Transfer
Acknowledged delivery of message from one DTN hop to the next and passing of reliability delivery responsibility
September 16, 2018
Anshul Kantawala

22. Custody Transfer (cont.)
Advantages
Relieves potentially resource-poor end nodes from maintaining end-to-end connection states
Useful for overcoming high loss rates along the delivery path
As reliable as typical end-to-end reliability
September 16, 2018
Anshul Kantawala

23. Protocol Translation and Convergence Layers
Bundle forwarding function assumes underlying reliable delivery capability with message boundaries
Convergence layer augments underlying network protocols appropriately
September 16, 2018
Anshul Kantawala

24. Time Synchronization Need for time synchronization
Provide a mechanism to deliver pre-programmed control instructions to be executed at future points in time
Use for scheduling, path selection and to remove expired pending messages
Propose time synchronization on the order of 1 ms
September 16, 2018
Anshul Kantawala

25. Security Each message contains Use public key cryptography
Identity of sender
Requested class of service (CoS)
Use public key cryptography
First DTN router verifies user and validates CoS request
Re-signs message using its key
Core routers need only cache keys of their neighbours
September 16, 2018
Anshul Kantawala

26. Congestion and Flow Control
Flow control is hop-by-hop
Uses underlying protocols mechanisms if they exist
Congestion control
Refers to contention of persistent storage at a DTN forwarder
Current approach uses a priority queue
Priority inversion and head-of-line blocking can occur
September 16, 2018
Anshul Kantawala

27. Application Interface
Applications must be able to operate in a regime where request/response time may exceed the longevity of the client and server processes
Application interface is non-blocking
Also has registration and callback functions between bundle-based applications and the local forwarding agent
September 16, 2018
Anshul Kantawala

28. Implementation
September 16, 2018
Anshul Kantawala

29. Implementation (cont.)
Prototype DTN system under Linux
Application interface
Rudimentary bundle forwarding across scheduled and “always on” connections
Detection of new and lost contacts
Two convergence layers
TCP/IP
Bundle-based proxy to the Berkeley mote network
September 16, 2018
Anshul Kantawala

30. Conclusion
DTN architecture attempts to provide interoperable communications between and among challenged networks
Design uses message switching with in-network retransmission, late-binding of names and routing tolerant of network partitioning
September 16, 2018
Anshul Kantawala