1. Disruption Tolerant Networking (DTN) An Architecture for Evolving Communications
Leigh Torgerson
Protocol Technology Lab Manager
Communications Architectures & Research (332)
Jet Propulsion Laboratory, California Institute of Technology
© 2019 California Institute of Technology. Government sponsorship acknowledged.

2. Presentation Topics: Why a Space Networking Architecture?
DTN History and Technical Overview
NASA Baseline Interplanetary Overlay Network (ION) DTN Suite
Current NASA Infusion Activities
JPL 332 Protocol Technology Lab Space Networking Testbed and Infusion Support for JPL
Questions
© 2019 California Institute of Technology. Government sponsorship acknowledged.

3. Why Networking?
Primarily because ops team management of point-to-point links doesn’t scale to the future Solar System Internetworking envisioned by NASA / CCSDS!
© 2019 California Institute of Technology. Government sponsorship acknowledged.
(15 Jan 2019)
REF: “Operations Concept for a Solar System Internetwork”, Interagency Operations Advisory Group (IOAG), 15 Oct, 2010.

4. Why Not Just Use Terrestrial Internet Protocols?
Six Key Assumptions Internet Protocol Design:
Networks are Richly Connected
Networks have Short Delay
Data Links are Symmetric and Bidirectional
Links have Low Error Rates, and that loss is due to Congestion
Network Nodes are Trustworthy
Connections are END-TO-END
Many link paths between routers; Links are persistent, symmetric and transitive
IP networks have end-to-end transit times on the order of milliseconds
IP networks assume that forward and return bandwidths are approximately the same and that full-duplex operation is normal
Error rates typically << 1e-8; TCP loss correction involves retransmission AND reduction of data throughput to reduce assumed congestion along the route
IP protocols were initially designed to assume that the network it is trustworthy. https and IPSEC provide additional security, but core protocols assume a trusted environment.
© 2019 California Institute of Technology. Government sponsorship acknowledged.

5. Network Characteristics We Face
Sparse network
Asymmetric; half-duplex or unidirectional links
Loss due to corruption, not congestion
Loss due to orbital geometry, antenna pointing errors, etc.
Inordinately long propagation times between nodes
Insecure – the threat: you don’t need a 34m dish to hack lunar comm links for instance
TCP breaks, inefficient
(OWLT in deep space) - ”chatty” protocols don’t work
© 2019 California Institute of Technology. Government sponsorship acknowledged.

6. Quick DTN Intro Video https://youtu.be/c_tcC51PIuM
Please see YouTube version
© 2019 California Institute of Technology. Government sponsorship acknowledged.

7. Brief Historical Outline
The Beginning:
1998 – A few members of JPL Space Mission Operations Standardization Program met with Dr. Vint Cerf to explain our work on enabling FTP/TCP/IP/IPSEC to adequately perform over geostationary satellites hops.
We knew that wouldn’t scale to interplanetary distances, and that more research was needed to really expand the internet into space.
Vint arranged DARPA funding and became a member of the team to study how to extend the internet into the solar system.
© 2019 California Institute of Technology. Government sponsorship acknowledged.

8. 1999: the IPN Architecture Core Team
Vint Cerf
Adrian Hooke
Scott Burleigh
Bob Durst
Keith Scott
Eric Travis
Leigh Torgerson
Howie Weiss
JPL:
Vint Cerf, Sr. Vice President of Internet Architecture, Worldcom Inc. and JPL Distinguished Visiting Scientist
Adrian Hooke, Scott Burleigh, Leigh Torgerson
MITRE:
Bob Durst, Keith Scott
Industry:
Eric Travis, Global Science and Technology
Howard Weiss, Sparta
Advisors:
Bob Braden (USC); Len Kleinrock (UCLA); Dave Mills (UDEL); Deborah Estrin (UCLA); Joe Touch (USC); John Wroclawski (MIT); Jon Crowcroft (UCL); John Klensin (ATT); and others
© 2019 California Institute of Technology. Government sponsorship acknowledged.
Salient point: initial interplanetary internet study and designs involved JPL, industry and academia, including pioneers of the internet.
RIP Adrian Hooke, Eric Travis

9. Beginnings: The Interplanetary Internet
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
RFC-5050
BUNDLE PROTOCOL
Title Interplanetary Internet (IPN)
Architectural Definition
Author(s) V. Cerf et al.
Filename draft-irtf-ipnrg-arch-00.txt
Pages 58
Date 18-May-01
This document describes the Interplanetary Internet - a communication system to provide Internet-like services across interplanetary distances in support of deep space exploration. Our approach, which we refer to as bundling, builds a store-and-forward overlay network above the transport layers of underlying networks. Bundling uses many of the techniques of electronic mail, but is directed toward interprocess communication, and is designed to operate in environments that have very long speed-of-light delays.
A URL for this Internet-Draft is
http//
© 2019 California Institute of Technology. Government sponsorship acknowledged.
Development phase – invent, build advocacy, standardize, try out on a deep space mission

10. Delay/Disruption-Tolerant Networking
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
JAXA DRTS
ISS Missions
ECOSTRESS
CAL
+ ~38 others
ISS – SCaN TESTBED
LADEE/LLCD
INFUSION PHASE
SSI now mainstream within CCSDS
DTN flight tested on Epoxi, and in LEO on EO-1 mission
DTN on optical links on Lunar Laser Comm Demonstration (LLCD) on the Lunar Atmosphere and Dust Environment Explorer (LADEE)
Commercial payloads on the ISS – widespread deployment including gateways on ISS and at HOSC
ESA and JAXA initiate experiments on ISS; JAXA plans DTN on next their lunar mission
Nation-wide DTN Engineering Network set up including NASA Centers, Universities in US and DLR in Germany
Core DTN protocols now standardized with IETF and CCSDS
© 2019 California Institute of Technology. Government sponsorship acknowledged.
DTN on Lunar Gateway

11. DTN Design Elements
Don’t replace terrestrial networking or legacy comm link technology
OVERLAY NETWORK!
Don’t reinvent reliable link technology
TCP/IP on land, Prox-1 for orbiter-to-surface
Deep Space Links needed a reliable link – so Licklider Transmission Protocol (LTP) was developed
Deal with loss and delay since contiguous end-to-end paths may be rare
Provide a store-and-forward mechanism at each node in the network
Fix what the IP design missed 40+ years ago
Build in security and integrity
Provide for Quality of Service
Deal with different routing needs of underlying networks
Pre-planned contact plans for space networks, opportunistic routing for ad-hoc situations, IP routing on the ground.
Design in provisions for network management, security key management, reliable multicast, projected data transfer types, etc.
Asynchronous Management Protocol (DTN-AMP), Delay-Tolerant Key Administration (DTKA), reliable multicast, special protocols for streaming video and applications that need in-order delivery
and make intermediate nodes secure by unencrypting only at the destination
© 2019 California Institute of Technology. Government sponsorship acknowledged.

12. DTN Stack Overview
Non-volatile storage for spacecraft allows DTN state and queued bundles to survive spacecraft reboots. Doesn’t have to be NVS; most any storage system will work.
© 2019 California Institute of Technology. Government sponsorship acknowledged.

13. https://youtu.be/cTZQJX5C4Ws
DTN for Human Space Flight - iPAS DTN 2019
Using Integrated Power, Avionics and Software (iPAS) distributed testbed between JPL and JSC.
Please see YouTube video at
Gateway / iPAS JSC/JPL DTN Use
© 2019 California Institute of Technology. Government sponsorship acknowledged.

14. Benefits of DTN Improved Operations: Interoperability and Reuse:
Store-and-forward mechanism
Automatic retransmission
Contact-based routing
Interoperability and Reuse:
CCSDS-standardized DTN protocol  International interoperability / reuse
Space Link Efficiency and Robustness:
Reliable links and automatic retransmissions  link efficiency
Error-corrected data throughput  more usable bandwidth
DTN Contact Graph Routing  multiple network paths and automation of routing of data.
Security:
Bundle Security Protocol (BSP)  end-to-end encryption and integrity checks.
Quality of Service:
256 priority levels which can be set for each individual data item
Allows sharing of links with a minimum of mission control management.
provides better data throughput than conventional methods, less complex scheduling and improved on-board resource utility.
© 2019 California Institute of Technology. Government sponsorship acknowledged.

15. Loss Recovery Loops -- Legacy vs DTN
Of interest to science community – lower latency, more efficient use of on-board storage
© 2019 California Institute of Technology. Government sponsorship acknowledged.

16. What DTN Software is Available?
Many public implementations available – DTN2 for university research and terrestrial uses; DTN also available in Java, C, Python – even for Android.
For space use, ION is the NASA standard
Interplanetary Overlay Network
Available on SourceForge
Contents: BP
LTP
Various convergence layers for LTP, TCP/IP, UDP/IP
Applications
AMS, BSSP, CFDP, DTPC
Performance & Network Monitoring
DTN-AMP
NASA provides encryption modules, convergence layers for TM/TC/AOS
© 2019 California Institute of Technology. Government sponsorship acknowledged.

17. Current Programs Using DTN 2019+
ISS
Ops side ( / files)
Payload ops – either direct or using ISS DTN Gateway
(ECOSTRESS, CAL, and about 38 other experiment packages)
GSFC PACE mission (FPGA, high-speed downlink implementation)
CubeSats (implementing DTN in the IRIS radio) – Lunar IceCube and others
KARI lunar program
HEOMD baseline for Lunar Gateway; commencing testing this year
© 2019 California Institute of Technology. Government sponsorship acknowledged.

18. AES Project DTN Deployment Plan
FY18 – Operational Usage
FY19 – Increasing Capabilities
FY20+ – Widespread Usage
ION 3.6.1
ION 3.6.2
ION 3.6.3
ION 3.6.4
NASA ION COMPLETE
Android port
Windows installer
SBSP update
Features added to RTEMS port
CGR SABR update
AMP Implementation Security Policy update
Bug fixes
IPV6 port
“Sticky” Routing
OCGR
Contact History Exch
Scalability
Node Auto-config
LTP Simplifications
Contact plan maturity
Bundle deliv times
Node Auto-configuration
Congestion Forecasting
High-Rate Support
Information Centric Networking
ISS
DTN GW H/W Upgrade, ION 3.6.1
TReK Demo, HRF, ECOSTRESS
Arcturus
ENCODE
AES
Avionics & SW Integration including CFS
Gateway Partners
EM-1 Cubesats
Gateway
NextSTEP Habitat
SCaN
Future Architecture Studies
High-Rate DTN Development
Mission & Infrastructure Infusion
NASA Infusion
Advocacy
HEO/SMD Study
Incorporate DTN into Ground and Flight Segments (RF and Optical)
Significant Operational Use
From Brenda Lyons’ DTN Team Presentation February 2019
KARI
Ground Testing, Consulting, Payload Planning
KPLO Support
Commercial Infusion
Industry Engagement, SBIR
Establish
Pilots
IETF Standards
Commercial Products
OGA Missions
R&D/Proof of Concept
Architecture Inclusion
Widespread Use of DTN
Opportunity Identification
Architecture Consultation

19. Infusion Support to Space Missions
Core DTN Development Team at JPL can provide resources for:
System Engineering - Rapid prototyping of multiple node (spacecraft or ground) scenarios with docker containers with actual DTN software
ION incorporation in flight software and radios,
End-to-end communications and data system system analysis, modeling, simulation and prototyping
Testing of implementations from Mission Ops to Spacecraft with AMMOS and DSN included
Future RF-in-the-loop testing with DTF-21
© 2019 California Institute of Technology. Government sponsorship acknowledged.

20. PTL EEIS Gold Standard Testbed
© 2019 California Institute of Technology. Government sponsorship acknowledged.

21. For More Information Associated with the Internet Society:
NASA DTN Site

22. Thank you for attending! Questions?

23. BACKUP SLIDES

24. The DTN “Bundle” DTN Bundle and Bundle Agent Functions
e2e Applications
(e.g., CFDP, CCSDS Packet, VOIP, Video)
Bundle API
Bundle Fragmentation & Reassembly
DTN Bundle and
Bundle Agent Functions
Bundle
Custody Transfer
Bundle
end-end Integrity
Bundle
Encryption
Security Key Management
Bundle
Expiration
Bundle Agent
Management Services
© 2019 California Institute of Technology. Government sponsorship acknowledged.
Bundle Routing
Convergence Layer (specific adapters that map Bundles to
underlying transmission services)
LTP
CCSDS- Space Packet
or Encap Packet
UDP/IP
TCP/IP
CCSDS
TM/TC/AOS
CCSDS
Proximity Link
SONET
Ethernet

25. Legacy Comm Protocols versus DTN
DTN is an IMPROVEMENT to current methods, not a replacement
© 2019 California Institute of Technology. Government sponsorship acknowledged.
CFDP = CCSDS File Delivery Protocol
DTN BP and LTP improve CFDP routing and reliability, automating timer settings
DTN/LTP automates multiplexing and prioritization of file and stream data
LTP provides for optional reliable stream data, with or without DTN bundle protocol
Same use of CFDP means DTN is transparent to end user or FSW file creation ops
L. Torgerson – 8/24/11

26. System Engineering Realities
© 2019 California Institute of Technology. Government sponsorship acknowledged.
Spending money in Phase C/D to save money/improve Mission Ops in Phase E is difficult because of the fact that Mission Ops isn’t usually funded until after launch, so no early investment by Mission Ops managers is possible.  DTN should be specified and planned for in Phases A & B! Engage Mission Ops and Science Teams as early as possible.