1 / 21

Real-Time and Store-and-Forward Delivery of Unmanned Airborne Vehicle Sensor Data

Real-Time and Store-and-Forward Delivery of Unmanned Airborne Vehicle Sensor Data. PI: Will Ivancic/GRC Co-PI: Don Sullivan/ARC . Approach

quant
Télécharger la présentation

Real-Time and Store-and-Forward Delivery of Unmanned Airborne Vehicle Sensor Data

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. 10 Minute Loop of 18 Slides

  2. Real-Time and Store-and-Forward Delivery of Unmanned Airborne Vehicle Sensor Data PI: Will Ivancic/GRC Co-PI: Don Sullivan/ARC

  3. Approach Collaborate with ARC UAV team and its satellite communications service providers to develop requirements and deploy advanced bandwidth efficient, reliable file transport protocols for the Global Hawk UAV Collaborate with appropriate router and radio manufacturers to develop a rate-based implementation of Saratoga and a modem link-property advertisement protocol Conduct integrated tests of the architecture and protocols using flight sensor data as a part of Global Hawk flight campaigns Real-Time and Store-and-Forward Delivery of Unmanned Airborne Vehicle (UAV) Sensor Data PI: Will Ivancic/GRC • Objectives • Develop and deploy a scalable communication architecture for NASA’s unmanned airborne vehicles (UAVs) based on Internet technologies • Improve the data throughput by developing and deploying technologies that enable the efficient use of the available communications links. Such technologies may include: • Delay/Disruption Tolerant Networking (DTN) or other store-carry-and forward techniques • Improvements to the Saratoga transport protocol (implementing a rate-based feature and congestion control) • Development of a protocol that advertises link properties from modem to router Global Hawk Command and Control Network Key Milestones Develop UAV communications architecture 12/09 Rate-based transport protocol initial deployment 2/10 Rate-based Saratoga Version 1 for single hop store and forward 6/11 Develop radio-to-router Layer-2 trigger protocol 5/12 Conduct integrated flight demonstration Summer 2012 TRLin= 4 TRLcurrent = 6 Co-PIs/Partners: Don Sullivan, NASA ARC

  4. Saratoga version 1 implementations • C (Charles Smith under contract to Cisco Systems) • Implementation licensed to CSIRO by Cisco. • Built for Speed (Raw I/O). • Streaming to be implemented in FPGA, File transfer may be implemented in FPGA. • C (Surrey Satellite Technology Limited – SSTL) • Implemented in Real-Time Operating System for high-speed image transfers from Low Earth Orbiting (LEO) satellites over highly asymmetric links. • PERL (NASA Glenn Research Center) • Sequential file transfer and rate-limiting implemented. • TCP Friendly Rate Control (TFRC) implementation patch by Kerrin Pine BEngMSc, Research Fellow Bio-Medical Physics, University of Aberdeen, Scotland • C++ (NASA Glenn Research Center) • Discovery, multiplexed file transfer, hooks for bundling and streaming and rate-limiting to be implemented. • Wireshark Dissector (Charles Smith) • http://sourceforge.net/projects/saratoga/files/

  5. Layer-2 Triggers

  6. Smart Modems • Modem's transmitting and receiving link rates can be varied over time due to the following: • Adaptive coding • Changes in Modulation to suit the channel characteristics. • Changes in transmission rate to suit the channel characteristics • Rate mismatch between RF link local area network. • Serial connections are less of a problem as clocks can be controlled by modem (at least the receiving clock) • Ethernet connections are becoming standard and result in rate mismatch between the LAN interface and the RF link. Ethernet 100 Mbps Ethernet 1 Gbps RF 3 Mbps Application Modem

  7. Issue / Problem • To condition traffic and get the most out of the modem's link capacity, applications need to know the modem's link conditions. • Figure 1 corresponds to existing commercial imaging satellites • Figure 2 is more generic • Desire is to have a standard method (protocol) for the application to understand the link conditions and adjust • Link Up/Down • Link Unreliable • Data Rates Serial Link RF 3 Mbps Application Modem Figure 1 Ethernet 100 Mbps Ethernet 1 Gbps RF 3 Mbps Application Modem Figure 2

  8. Status (page 1) • Submitted Internet Draft to Mobile Operations Research Group (mobopts) • Modem Link Properties Advertisement Protocol (draft-ivancic-mobopts-modemlpa-01 Updated April 2012, Expires October 2012 • Although the draft uses modems as the device between some network cloud and the upstream network attached devices, a cryptographic system in place of the modem has very similar issues that the protocol handles. • Uses link-local multicast, unicast and currently site-local (ipv6) and organizationally scoped (ipv4) multicast. There are pros and cons to each approach • Proposed solution allows applications and concepts such as disruption/delay tolerant networking (DTN) to have some clue of network connectivity and link layer characteristics even if the host / router / forwarding agent is multiple hops away from the modem.

  9. Status (page 2) • The authors of modemLPA have been in discussion with the authors of "Dynamic Link Exchange Protocol (DLEP) to determine if DLEP will fulfill the needs that ModemLPA is targeted at. • DLEP is currently a client/server session oriented protocol that provides link layer information to directly connected • It should be possible to use the DLEP message formats without all the signaling required for the DLEP client/server session to perform the functions addressed in this draft. • The modem would simply provide link states out via multicast or unicast UDP datagrams (DLEP-Lite). Whether or not this is a good idea, or acceptable to the manet group, is yet to be determined. • Paper published and presented at IEEE Aerospace Conference, March 2012.

  10. 3 Year Summary / Accomplishments

  11. Experiment Operations • Obtained a very good understanding of the NASA (and DOD) Global Hawk communications system • Obtained a very good understanding of how the Experiments are run • Attended Hurricane Karl experiment • Talked with mission manager • Talked with researchers to understand communications needs and how the control their instruments • Obtained a excellent understanding of the interaction between Global Hawk pilots and Researchers.

  12. Global Hawk Control Room at Dryden Obtained

  13. NASA Global Hawk Communication Network(most complex scenario proposed) Ku Band Satellite - B Ku Band Satellite - A No Network Mobility and Single Hop therefore: No need for DTN or Mobile Networking > 3 Mbps Bidirectional Link NASA Dryden Disconnection During Satellite Handover Due to Repointing L3-Com Ku-Band Terminal

  14. Evaluation of Reliable Rate-Based Protocols • Desire to evaluate rate-based protocols for high-speed data delivery • Saratoga version 1 • Negative Acknowledgement (NACK) - Oriented Reliable Multicast (NORM)Transport Protocol • CCSDS File Delivery Protocol (CFDP) – Class 2 • CFDP – Class 1 over DTN over Licklider Transmission Protocol (LTP) over IP • Limited Funds and FTE so moved to DTN project under SCaN/SOMD • Results/Conclusions • Saratoga PERL performed as expected at line rate • NORM worked out of the box • Performance was consistent across a wide variety of link conditions • Multicast-oriented design best suits those particular environments where data must be transferred to a large amount of receiver nodes at once • CCSDS Protocols were difficult to get code for • LTP used Trinity College of Dublin code • Buggy and performed poorly due to implementation • CFP code obtained from GSFC • Performed well

  15. Rate-Based Protocol Testing Results

  16. Protocol Enhancing Proxies (PEPs) Testbed Tested all of these protocols with and without a PEP for delays of 0, 100, 300 and 600 millisecond and bit-error-rates of 0, 10-7 and 10-5

  17. PEP Conclusions • A NACK-based file transfer protocol such as Saratoga will out perform a TCP-base file protocols that use modern TCP implementations or a PEP. This was the expected result. • A PEP designed to improve TCP performance over large Bandwidth Delay Product (BDP) link will not improve interactive communications of single packet transfers. This also was the expected result. • For our particular system, a high BDP link with no competing traffic and very few errors, the self-tuning capabilities of modern TCP implementation provide nearly identical performance to deployment of a PEP and require no configuration or tuning. This was a bit of a surprise. • The SCPS PEP (an most if not all other PEPs) must be configured for the BDP characteristics of the link they are compensating for. If the link BDP changes, the PEP configuration must also be updated. • Paper published in the • March 2012 IEEE Aerospace Conference proceedings

  18. Presentationsand Publications

  19. Publications and Presentations • William D. Ivancic, Donald V. Sullivan: "Delivery of Unmanned Aerial Vehicle Data," Earth Science Technology Forum 2010, Crystal City, VA, June 22-24, 2010 • William D. Ivancic, Donald V. Sullivan, David Stewart, Patrick E. Finch: “An Evaluation of Protocols for UAV Science Applications”, Earth Science Technology Forum 2011, Pasadena, CA, June 21-23, 2011 • L. Wood, W. Eddy, C. Smith, W. Ivancic, C. Jackson: “Taking Saratoga from Space-Based Ground Sensors to Ground-Based Space Sensors”, IEEE Aerospace Conference 2011 • A. Shahriar, M. Atiquzzaman, L. Wood, W. Ivancic: “A Sender-based TFRC for Saratoga: A Rate Control Mechanism for a Space-Friendly Transfer Protocol”, IEEE Aerospace Conference 2011 • W. Ivancic, D. Stewart: “Advanced Networks in Motion Mobile Sensorweb” (Documenting work done in previous AIST task), IEEE Aerospace Conference 2011 • W. Ivancic: “Applying Web-Based Tools for Research, Engineering and Operations”, IEEE Aerospace Conference 2011 • W. Ivancic, L. Wood, R. Asati, D. Floreani, D. Shell: “Modem Link-Property Advertisements”, IEEE Aerospace Conference 2012 • P. Finch, D. Sullivan, W. Ivancic: “An Evaluation of Protocol Enhancing Proxies and Modern File Transport Protocols for Geostationary Satellite Communication”, IEEE Aerospace Conference 2012

  20. Internet Drafts – work in progress • L. Wood, W. Eddy, C. Smith, W. Ivancic, C. Jackson: “Saratoga: A Scalable File Transfer Protocol,” draft-wood-tsvwg-saratoga-11,work in progress, updated April 2012, Expires October 2012 • L. Wood, W. Eddy, W. Ivancic: “Congestion control for the Saratoga protocol” (draft-wood-tsvwg-saratoga-congestion-control-01), work in progress, updated April 2012, Expires October 2012 • W. Eddy, L. Wood, W. Ivancic: “TFRC-based Congestion Control for Saratoga” (draft-eddy-tsvwg-saratoga-tfrc-01), work in progress, updated April 2012, Expires October 2012 • W. Ivancic, L. Wood, R. Asati, D. Floreani, D. Shell: “Modem Link Properties Advertisement Protocol” (draft-ivancic-mobopts-modemlpa-01), work in progress, Updated April 2012, Expires October 2012

  21. Programmatics • 3 Year effort • $350K per year to GRC • 1 WYE (Contractor) and 1/2 FTE (Civil Servant) • Protocol Research • $150K per year to Ames Research Center • Command and Control of Payloads. • June 2009 – May 2012 • Effort completed

More Related