Analysis of Network Outage Reports1
Cybersecurity and Communications Reliability Division – Public Safety & Homeland Security Bureau
John Healy
December 2015
Frequency of Outages by Month2
Standards for Outage Frequencies
Took average of January 2012 through December 2012 as the standard
3
Outage Index – For All Outages Except DS3-Simplex
4
See ATIS-0100021 Analysis of FCC-Reportable Service Outage Data for a description of the method used to calculate the outage index.
Procedural Events by Source5
New Planned Maintenance Cause Codes
6
DS3 (Non-Simplex) Events7
Over the Last 4 Months• Hardware Failures
are up• Power Failures are
up• Cable Damage
events are up
Big DS3 Outages8
DS3-Simplex9
Fiber Cut10
E91111
Very Large E911 Outages12
E911 – Phase 2 Events13
Phase 2 vs. All Other E911 Outages14
Wireline15
Wireline outages are all outages for which the Reason Reportable was at least 900,000 wireline user-minutes.
Wireless16
VoIP Reporting17
Environmental – External18
SS7 19
Malicious Events20
Malicious Activity21
*Please note the total for 2015 is thru September.
Year Vandalism Cable Theft Gun ShotsTotal Malicious
Events
2012 35% (83) 8% (18) 7% (15) 207
2013 37% (83) 4% (9) 9% (22) 224
2014 51% (99) 6% (11) 11% (22) 193
2015 42% (75) 7% (13) 12% (21) 180*
Duration Trends: DS3 Non-Simplex
22
DS3 Non-Simplex Outages are statistically shorter in 2015 compared with 2014 (using a Mann-Whitney test)
Duration Trends: E911
23
E911 Outages are statistically longer in 2015 compared with 2014 (using a Mann-Whitney test)
Duration Trends: Wireline
24
The duration of wireline outages is statistically unchanged from 2014 to 2015 (using a Mann-Whitney test)
Duration Trends: Wireless
25
The duration of wireless outages is statistically unchanged from 2014 to 2015 (using a Mann-Whitney test)
Test for Trend in Event Duration
26
Non-DS3-Simplex OutagesData from 1/12 thru 9/15
Conclusion: Non DS3-Simplex Outages are unchanged in duration from 1/12 to 9/15
DS3-Simplex Events OnlyData from 1/12 thru 9/15
Conclusion: DS3-Simplex Events are unchanged in duration from 1/12 to 9/15
Rank Correlation of duration with date is -.010 and is not significant at the .01 level
Rank Correlation of duration with date is -.011 and is not significant at the .05 level
911 Certification
Preparing a public report with the non-confidential information
Having discussions with companies about their filings
Comments and suggestions are greatly appreciated
27
NRSC Activities
Current Teams Best Practices Team Outage Reporting Advisory Team
Improvements to the NORS System Improvements to the template Recommended ways to communicate information Improvements to NORS/DIRS manuals
IP Transition Team Consistent Use of Cause Codes
28
Top Related