Transcript mirk-minwara-may12

Material Identification Reflectivity
Kernel (MIRK) for
MCM/Mining Single Pass Detect-toEngage (DTE) Operations
Radm John Pearson, USN (Ret)
9 MAY 2012
Prometheus Inc.
1
DTE Challenges/Unmet Needs
• Sonar detection/classification currently/usually
only at “possible/Mine-like object” level
– Too many objects of interest for many scenarios
– Too many false targets in cluttered environments like the
Persian Gulf
– Too many neutralization weapons required/wasted due
to false targets
• Current systems post-mission analysis (PMA) time
precludes DTE
• Successful discrimination and neutralization
– Successful discrimination currently questionable
– True autonomous DTE operations not available
Prometheus Inc.
2
Background: What is MIRK?
• Proven USAF Program of Record
– Three phases of SBIR development and transition
– Material Identification-Synthetic Aperture Radar (MISAR)
• Demonstrated sonar capability
– Phase I/II SBIR’s with Mk 48 submarine torpedo program
– Initial NUWC analysis after MIRK processing of torpedo
sonar data showed highly enhanced performance
– Initial MIRK processing of real data shows promising
results on two other sonars
• BQQ-10 submarine active sail array
• SQS-53C surface ship sonar
Prometheus Inc.
3
4
Phase I SBIR Hypotheses/Approach
Echo returns from active interrogation of an
underwater object contain reflectivity kernel (RK)
clues
• Deconvolution - Not a new problem/Knowing
transmitted signal and echo return, solve for RK
• New, unique mathematical technique devised by Prometheus
 Use Time vs. Frequency Domain approach
 Highly stable, real-time processing
• Approach: Classify active sonar contacts based on material
identification technique
– Formulate theoretical hypotheses showing the existence
of an RK in active sonar echo data
– Develop algorithm to extract the RK from an actual
received echo, given parameters of transmitted sonar
signal
Prometheus Inc.
5
Prometheus Inc.
The Inverse Problem
Given the input signal (s) and the scattered or
output signal (f), find the Reflectivity Kernel (K)
f (t ) 
Input


K
(
t
,
t
)
s
(
t
)dt ' noise

Kernel
Output
Prometheus Inc.
Demonstrated and Potential Capabilities
• Demonstrations using Mk 48 torpedo at sea data
– Processing of torpedo data recorded during SOBA range
exercises containing the WSTTT target showed significant
detection/classification performance enhancement.
• Demonstrated over 95% reduction in false alarm rate
over the torpedo performance
• NOT aspect or elevation angle dependent
• Algorithm computational efficiency allows integration
into torpedo software in the near term
– Algorithm applied to 53C data has shown similar capabilities.
Ongoing efforts will quantify performance enhancements.
– Potential to detect/classify/identify (?) mines, bottomed targets
and underwater vehicles/satisfy CNO Abbreviated Acquisition
Program (AAP) requirement for improved Pdc, Pid and Pfa in
highly complex environments (enable sonar target discrimination
capability in DTE operations)
Prometheus Inc.
6
7
MIRK Detection / Classification
Rocks & Rock Ridges
Target Detection / Classification
Prometheus Inc.
Road to Effective MIRK DTE
Operations
• Short term
– Demonstrate MCM proof of concept using current MIRK
target discrimination capability with recorded MCM
sonar data
• Show significant reduction in false alarm rate in highly cluttered
environment
• Software upgrade only
– Investigate MIRK type processing for laser mine detection
sensors
• Long term
– MIRK processing in all MCM sonars to enable real time
DTE
– MIRK processing for future mining target detection
devices (TDD) using acoustic detection/homing
Prometheus Inc.
8
Advantages
•
•
•
•
•
Relatively insensitive to operating frequency
Current waveforms provide adequate bandwidths
Robust in low signal to noise ratio (SNR) echo returns
Operates in parallel with existing detect/classify functions
Can be inserted into many active sonar systems with only a
software impact
– Torpedoes: MK-54, CBASS, MK-48 (Current Phase II SBIR)
– Submarine: AN/BQQ-10 (Acoustic Rapid Commercial-Off-The-Shelf
Insertion (ARCI))
– Surface ship: AN/SQS-53C
– Littoral and Mine Warfare: MCM-1 class ships’ SQQ-32 HFWB (ARCI)
sonar, Littoral Combat Ship Mission package, UUVs
MIRK is a single ping detection/classification
algorithm that operates on active sonar returns
over their operational listening range without
need of imaging or multiple looks
Prometheus Inc.
9
10
Recommendations
• Short Term
– Labs/MCM contractors provide MCM sonar data for MIRK
processing (e.g. SQQ-32 HFWB (ARCI), EOD UUV, AN/AQS20/24, AMNS, etc. sonar)
– Prometheus apply MIRK processing, labs conduct analysis
of processing results for validation
– Similar action for laser mine detectors
– Similar action for additional ASW sonar processing and
validation
• Long Term
– Refine and optimize MIRK processing for MCM sonars
– Fully integrate MIRK into current and planned MCM sonars
– Evaluate MIRK DTE operations in future mining TDD’s
Prometheus Inc.