B. IMPACT OF CLUTTER ON SENSOR PRFORMANCE IN LAND COMBAT
The Longbow program was further along than any others; the Longbow approach
used information provided by its high range resolution radar return from the stationary
target as opposed to a Doppler-shifted return from a moving target.
As a follow-on to the work modeling radar clutter for engagement of air targets,
STD got a study from the Army’s Model Improvement and Simulation Management
Agency to look at how clutter affected engagement of stationary ground targets. The
Longbow program was at that time further along than any other at trying to exploit highrange-
resolution radar signals to do detection and identification of stationary targets. It
was selected as the technology development on which to focus.
Tactical radars have low resolution in azimuth and elevation, limited by antenna
size. The resolution in range is limited by bandwidth and can be much higher. “High
range resolution” means that the radar “pixel” size (in the downrange direction) is small
compared with targets of interest. For a Longbow-sized antenna operating in K-Band,
with a 1 GHz bandwidth, the radar could distinguish “spots” on the ground about 100 m
wide and 15 cm deep. Typically, one would group these range cells, looking for “hot
spots,” then process the individual cells. A ground patch of about 30 range resolution
cells (even numbered cells in blue, odd numbered clear) with a tank in the middle might
The task for the clutter-suppression algorithm is to use the 30 or so returns, called
a “range profile,” from the wide, thin ground patches to determine whether a target of
interest is present. This is very challenging for a number of reasons:
• Each range cell has much more nontarget area than target area. Therefore, the
return may not be dominated by the target.
• The target may overlap adjacent patches, either in range or in cross range.
• The target orientation is unknown.
• For a turreted vehicle, the turret chassis angle will be unknown.
• The returns will vary rapidly with viewing angle. For example, there is no
reason to believe there will be any similarities in target returns between the
situation below and the one above.
Note that in the application of the high-range resolution technique to the air-to-air
environment, all these challenges are substantially mitigated or nonexistent.
Experienced analysts with a strong technical background were able to intuit, on
first hearing a description of the Longbow program, that performance against stationary
targets would be its Achilles heel (1987). This was identified as a research area that STD
successfully pursued as part of its sensors and target acquisition focus. The research on
the Longbow algorithms revealed the approach to be incapable of delivering needed
operational performance (1992). This was stated publicly as part of the Bottom-up
Review (1993), as was the fact that a forward-looking infrared could provide adequate
target location for the RF missile, thus providing stationary target capability through a
different approach. In 1995 the IOT&E revealed the difficulties with detection and
classification in a formal test with publicized results. As late as the summer of 2000 we
were still discovering the shortcomings of those tests.
In 2003 as part of Iraqi freedom, Apache operators found:
• The system was effective against moving targets and in bad weather.
• False alarms were excessive, especially for stationary targets.
• Most operators lacked confidence in the fire control radar for target location
• The forward-looking infrared was the primary target acquisition sensor for
the RF Hellfire
• There were too many false SA-8 declarations.
How do we avoid spending money and delaying programs by pursuing features
that will not be achieved with the current approach, but are achievable in other ways?
How can we ensure that future operators are not surprised during an operation by
what had been intuited 16 years earlier, known with certainty 10 years earlier, and
revealed in tests 8 years earlier?
The Longbow Ka-band (MMW) for ground targets covers 90° in azimuth and quite narrow angle in elevation. The coverage of tank sized and combat relevant RCS signatures on battlefield does not exceed 8km under optimal conditions. Thus we calculate the field of coverage of ground targets by determining the area of sector it covers. 90° / 360 * pi*radius 8² = 50.27km², not counting in blind spot of 500-1000m right infront of the helicopter, depending on its altitude the blindspot is eleminated or extended.
The Longbow does feature air-to-air mode which was unsatisfactory (AH-64D Block2) not known in Block3, but the E suffix was given after the Longbow problems have been reduced/eleminated or at least decreased to some closer satisfactory level to come closer to specifications of the Longbow programm that were pursued.
Arbalet K,Ka-band (MMW broad) covers under optimal conditions 8 up 12km, while it is usually always closer to 8km rather then 12km. The azimuth is slightly higher with 120° and elevation is also higher since the Arbalet antenna can move in 2D inside the nose dome.
The Arbalet covers between 67km²(8km range) to 150km². The problems of this come similiar to that of Longbow in clutter distorting the recieved data to process a clear picture of the situation especially for stationary targets without distingtive RCS not to mention what mess it would be on a battlefield were tons of ammunition is fired into the field and remains of ammunition is increasing steadily the clutter the radars recieve.
Most of the time FLIR and intel from observering/commanding helicopters is more useful than using MMW radar.