| NSF Org: |
CNS Division Of Computer and Network Systems |
| Recipient: |
|
| Initial Amendment Date: | March 23, 2018 |
| Latest Amendment Date: | March 23, 2018 |
| Award Number: | 1806785 |
| Award Instrument: | Standard Grant |
| Program Manager: |
Darleen Fisher
CNS Division Of Computer and Network Systems CSE Directorate for Computer and Information Science and Engineering |
| Start Date: | April 1, 2018 |
| End Date: | March 31, 2019 (Estimated) |
| Total Intended Award Amount: | $198,850.00 |
| Total Awarded Amount to Date: | $198,850.00 |
| Funds Obligated to Date: |
|
| History of Investigator: |
|
| Recipient Sponsored Research Office: |
3720 S FLOWER ST FL 3 LOS ANGELES CA US 90033 (213)740-7762 |
| Sponsor Congressional District: |
|
| Primary Place of Performance: |
4676 Admiralty Way, Ste 1001 Marina del Rey CA US 90292-6601 |
| Primary Place of
Performance Congressional District: |
|
| Unique Entity Identifier (UEI): |
|
| Parent UEI: |
|
| NSF Program(s): | Networking Technology and Syst |
| Primary Program Source: |
|
| Program Reference Code(s): |
|
| Program Element Code(s): |
|
| Award Agency Code: | 4900 |
| Fund Agency Code: | 4900 |
| Assistance Listing Number(s): | 47.070 |
ABSTRACT
Natural disasters such as hurricanes and blizzards cause economic and social disruption in the U.S. and globally. First responders need information about the extent of problems, citizens would like information about cities where friends and relatives live, and government planners and researchers would like to design stronger infrastructure. However, information about the location of problems and speed of recovery can be slow to emerge during a disaster, and often remains imprecise and incomplete days or even weeks afterwards. Internet outage measurements can be a sensor to measure the effects natural disasters. The principal investigator and his team have shown that Internet outages can be observed from a few central sites--their work on Trinocular has been peer reviewed and they have been collecting data twenty four hours a day, seven days a week for more than two years. Serious natural disasters often result in Internet outages, typically because of power or communication loss due to utility pole failure, flooding, or wire breakage; this correlation has been shown by several groups. While the team makes outage data available to researchers today, this data is currently inaccessible to lay-people and even scientists, since they cannot easily browse the data or drill down into regions of interest. This project makes Internet outage data accessible through a new, interactive website, with goals of directly assisting citizens, first responders, and scientists to understand the scope of disasters and their recovery. Second, this website and data is used to compare the team's outage data with two sources of "ground truth" outages from disasters. A primary source of ground truth will be the Federal Communications Commission's (FCC's) NORS (Network Outage Reporting System): these industry-reported outages are the data the U.S. Government uses today to assess the status of the U.S. telecommunications system. This work will build on ongoing collaboration with the FCC, in which the team provides the FCC with its outage data and works with them to compare it with the FCC's proprietary data. The team also expects to compare its data to public reports of utility outages.
This work builds a better understanding of the relationship between detectable outages in the Internet and public utility outages. This understanding will result from making current data more accessible to researchers and the public, and through our comparisons of Trinocular to ground truth data sources. In addition, making existing outage data more accessible to other researchers. This work promotes a better and more timely understanding of the consequences of natural disasters and the speed of recovery.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
PROJECT OUTCOMES REPORT
Disclaimer
This Project Outcomes Report for the General Public is displayed verbatim as submitted by the Principal Investigator (PI) for this award. Any opinions, findings, and conclusions or recommendations expressed in this Report are those of the PI and do not necessarily reflect the views of the National Science Foundation; NSF has not approved or endorsed its content.
Interactive Internet Outage Visualization to Assess Disaster Recovery (IIOVADR) is a project supporting the use of Internet outage measurements to help understand and recover from natural disasters.
Internet outage measurements can be a sensor to measure the effects natural disasters. We have shown that we can observe Internet outages from a few central sites?the Trinocular outage detection system has been peer reviewed and we have been collecting data 24x7 for more than two years. Serious natural disasters often result in Internet outages, typically because of power or communications loss due to utility poll failure, flooding, or wire breakage; this correlation has been shown by several groups.
This project made four advancements advancing the intellectual merit and broader impact.
1. The project improved visualization of weather-related outages, adding near-real-time reporting of outages to the project's existing geographic website at .
The project built a new near-real-time data processing pipeline to gather outage data, stream it back to USC/ISI, integrate observations from 6 sites, geolocate the results, and feed them into a database. We extended the Outage Visualization website to track dynamic updates to that data. The results of this work are available to the public at https://outage.ant.isi.edu/.
2. The project improved support for interaction with the website. We improved the Outage Visualization website by incorporating data from Hurricanes from the NOAA website, and to visualize them in the outage processing system.
In applying this to existing events like Hurricane Harvey, this data revealed that the relationship between network outages and the hurricane were more complex than previously understood.
Using this website to examine Hurricane Harvey 2017 (Texas), we saw a complex relationship between Hurricane Harvey location and Internet outages over time. This event can be viewed on the Outage Visualization website via the link https://ant.isi.edu/url/harvey2017 with an image harvey_2017_screenshot.png.
Using the Outage Visualization website to examine Hurricane Irma 2017 (Florida), the data shows significant outages when Hurricane Irma passes over Florida, followed by rapid recovery over next 24 hours. This data is visible on the Outage Visualization website via the link https://ant.isi.edu/url/irma2017 and with an image irma_2017_screenshot.png.
Using the Outage Visualization website to examine Hurricane Maria 2017 (Puerto Rico), the results show significant outages when Hurricane Maria passes over Puerto Rico, followed by extended outages over following weeks due to difficulty restoring power and infrastructure. This event is in the Outage Visualization website at https://ant.isi.edu/url/maria2017 and in the image maria_2017_screenshot.png.
As a non-hurricane example, the project examined Internet outages in Venezuela in March 2019, presumably due to power outages related to economic problems. This event is on the Outage Visualization website via the link https://ant.isi.edu/url/str201903ve and in image venezuela_2019_screenshot.png.
3. The project Improved the science behind the visualization by comparing USC/ISI outage data with other data sources.
We worked with other researchers to compare USC/ISI outage work with other data sources. Because their data is not public, we cannot directly compare to their results. However, comparisons with published results was sufficient to guide design of new algorithms.
Project researchers designed new algorithms to address ?sparse blocks?, a class of networks where Trinocular frequently gives false outages, and to address ISP-based address renumbering. In addition, we provided a theoretical definition of outages that is independent of any measurement system, and showed that three observers are sufficient to converge on the true value. These new algorithms and results are described in the technical report ISI-TR-733 ?Improving the Optics of Active Outage Detection (extended)? by Guillermo Baltra and John Heidemann. (Also currently under peer review.)
4. The project distributed data to other researchers: Over the course of the project (April 2018 to March 2019), we provided 665 datasets (11.5TB after compression) to 27 researchers.
Last Modified: 06/10/2019
Modified by: John S Heidemann
Please report errors in award information by writing to: awardsearch@nsf.gov.
