Running Strategic Analytics Affecting Europe and Africa
Open PDFDocument date: -
Pages: 14
Genre: slides
Banner: TOP SECRET//SI//REL USA, FVEYS
Classification: TOP SECRET SI
Dissemination: REL USA, FVEY
This Top Secret presentation from the NSA ECC (European Cryptologic Center) outlines the strategic challenges and technical methodology for conducting signal intelligence operations across Europe and Africa. It emphasizes a critical reliance on "Second Party" partnerships, particularly with the UK's GCHQ, and details the use of mass surveillance tools like XKEYSCORE to track terrorism, smuggling, and human trafficking networks.
My notes
TOP SECRET//COMINT//REL USA, FVEYS
(U) Running Strategic Analytics Affecting Europe and Africa
Region: Europe, Middle East
(Israel), and Africa :
[REDACTED], ECC
UNITED STATES CYBER COMMAND
CENTRAL SECURITY SERVICE
NSA/CSS Europe and Africa
Regional Presence Global Reach
NATIONAL SECURITY AGENCY
UNITED STATES OF AMERICA
NSA/CSS EUROPE
WIESBADEN - GERMANY
The overall classification of this briefing is:
TOP SECRET//COMINT//REL USA, FVEYS//20291123
Outline
- (U) Background
- (U) Problem Definition & Challenge
- (U) Our AOR: Europe - Africa
- (U) Examples for Europe - Africa
- (U) Enrichment and Data Flow
- (U) Real-time, batch, XKEYSCORE
- (U) Conclusions
33
UNCLASSIFIED//FOR OFFICIAL USE ONLY
(U) Terrorists Transit via Europe
- (U) Communication
- Transit Points
- (U) Partners
- Second Party
- Third Party
- (U) Relationships
- EUCOM
- AFRICOM
- CENTCOM
NCEUR Support to EUCOM
(U) US needs partners for data & to help capture, confine....
CONFIDENTIAL//REL USA, FVEYS
34
(U) Challenge: Integrating Tactical & National Collection
- (C//FVEY) Collection with HF/VHF/UHF
- Digital packets
- Analog comms
- Noise issues, lack of experience with these types of signals
- (C//FVEY) Tactical versus National (Strategic) Collection
- RTRG
- DISTILLERY
37
CONFIDENTIAL//REL USA, FVEYS
(U) Analytics for Targets in Europe
- (C//FVEY) OPSEC Savvy Targets
- “...most terrorists stop thru Europe”
- (TS//FVEY) Use advanced techniques
- Steganography
- Forensics or Analytics on front end
- Encryption
- Takes time and has "black hole" issue
- Steganography
- (TS//SI//FVEY) Reliance on "special" collection
- GCHQ and FAA
- Problems processing w/r to TS
TOP SECRET//SI//REL USA, FVEYS
(U) Analytics for Identity Intelligence
| (U) Human Trafficking | (C//FVEY) Operations from Jordan to Syria in both directions; Sahel | Metadata for geolocation; content for confirmation |
| (U) Weapons Smuggling | (C//FVEY) From Libya to Sahel | Metadata for geolocation; content for confirmation |
| (U) Drug Smuggling | (C//FVEY) Sahel and financing of terrorism; Balkans into Europe | Metadata for geolocation; content for confirmation |
| (U) Biometrics & Elections | (C//FVEY) Used in Africa | Need collection assets |
CONFIDENTIAL//REL USA, FVEYS
41
(U) Enrichment Sources
- (U) Air Breather, HF & UHF/VHF
- (C//FVEY) Big Pipe & FORNSAT
- (U) Military SIGINT Services
- (U//FOUO) Forensics
- (U) Third Party Sources
- (C//FVEY) Second Party
- GCHQ is critical for mission
QRC Package
3rd Party Partner Sharing
Forensics
Computer Forensics
(C//FVEY) Key Idea: Low Priority of AFRICA may cause loss of metadata and content; makes "Discovery" more uncertain
CONFIDENTIAL//REL USA, FVEYS
(U) Enrichment: SIGDEV & GCHQ QFDs
Account Allocations by TOPI
(March 2012)
- S2A: 4%
- S2B: 0%
- S2C: 1%
- S2D: 3%
- S2E: 6%
- S2F: 2%
- V23: 1%
- V22: 1%
- Other: 12%
- FGS: 5%
- S2G: 6%
- S2H: 1%
- FTS: 8%
- FHS: 2%
- SSG: 1%
- S2I: 22%
- F22: 17%
- F6: 9%
- (S//FVEY) 54% of current ECC DNI tasking based on QFD data
- (S//FVEY) QFDs provide better access to metadata for European & North African targets than any other access at ECC due to poor passive collection
- (C//FVEY) Flexibility provided by the use of TDIs and the first stage query allows for better target discovery and development
Slide taken from ECC archives.
(C//FVEY) Much of ECC data comes from GCHQ QFDs
SECRET//REL USA, FVEYS
(U) Data Flow Integration is Constant Headache
Access
Signal Acquisition (RF or Optical) -> Signal Conditioning: Amplification, Distribution -> Receiver/Downconverter (RF) -> Signal Demodulation (RF)
Exploitation
Signal Demultiplexing -> Channel Processing -> Transport Metadata Capture -> Target Selection -> Voice/Fax/Data Processing and Recording
S2 TOPI/ECC
Target Development
Target Tasking
Translation/Transcription
Intelligence Reporting
Transition to the CLOUD
Data Mgmt
Events
Selectors
Voice/fax/data files
Whose job? S1, S3, T?
46
SECRET//REL USA, FVEYS
(U) "Real Time" Analytics
- (U) Nascent Analytics with unclear definition of "real time"
- How fast is alerting?
- (C//FVEY) DISTILLERY
- Pulled from GHOSTMACHINE stack
- (U) NIAGARAFILES
- File based
- Starting to gain experience
- (C//FVEY) RTRG
- Tools not integrated into ECC
- Data Sets are sparse
- Tactically oriented
- Unregulated alerts can quickly spam user
- (C//FVEY) ECC Current Effort:
- Focused on NTOC and Distributed Denial of Service attack alerting
- Uses DISTILLERY
(U) How fast is real time?
CONFIDENTIAL//REL USA, FVEYS
47
(U) Batch: MapReduce Analytics
- (U) Batch oriented versus streaming
- Run every 15 min to once a day or so
- Not streaming
- (U) Good Data Storage
- Good access outward to MDR-1, MDR-2
- Days to years of storage
- Promotion (?)
- (U) Complex Analytics like "Pattern of Life"
- Reasonable amount of processing cycles at the front end collection system (not yet tested)
- (U) Session can be quite long and still captured (not yet tested)
- (U) UUID's (identifying sessions) are workable
- (U) No experience yet sharing with second and third party partners
- (U) Unknown level of entry training required
- Menwith Hill has WHIZBANG
(C//FVEY) Batch gives you access to data 24 number
hours ago
UNCLASSIFIED//FOR OFFICIAL USE ONLY
48
(U) Xkeyscore Fingerprints
- (C//FVEY) Streaming
- Data available one hour later?
- Most do pulls up to yesterday
- (U) Good Data Storage
- RAW content: 3 days to a couple of weeks
- Metadata: 90+ days
- (U) Complex Analytics like "Pattern of Life"
- Reasonable amount of processing cycles at the front end collection system
- (U) Session can be quite long and still captured
- (U) UUID's are workable
- (U) Good for sharing with second and third party
- (U) Relatively low level of entry training required
(U) XKS fingerprints great for streaming
CONFIDENTIAL//REL USA, FVEYS
49
(U) Key Take Aways
- (U//FOUO) Discovery in Africa is based on "we do not know what we do not see"
- Unknown Unknown from url: https://wiki.nsa.ic.gov/wiki/NTOC-E_discovery_tradecraft
- (U) Europe has Opsec savvy CT targets
- (U) Analytics involve partners
- 3rd Party in future
- (U) Limited Resources: Processing Power & BW
50
UNCLASSIFIED//FOR OFFICIAL USE ONLY
NSA/CSS Europe & Africa
EUROPEAN CRYPTOLOGIC CENTER
NSA/CSS EUROPE AND AFRICA
QUESTIONS?
51
UNCLASSIFIED//FOR OFFICIAL USE ONLY