Aerospace and Electronic Systems Magazine September 2016 - 56

Proof-of-Concept Airborne Sense and Avoid System

Figure 3.

Intruder aircraft details.

in mixed equipage environment. GA-ASI also provided a second
Predator B (N308HK) to demonstrate the first ever UAS vs. UAS
collision avoidance capability. Figure 3 below describes the details
on relevant equipage and tests for each intruder aircraft.

SURVEILLANCE SENSORS AND STM PERFORMANCE
FAA STM USING DISCRETE SURVEILLANCE SOURCES
Two variants of the FAA STM were tested, each using a discrete surveillance source to investigate the utility of that surveillance source for collision avoidance. The ADS-B STM variant
was evaluated for CA against cooperative intruders and likely,
the ATAR STM variant against noncooperative intruders. ADSB data transmitted by participation aircraft was recorded using
the ACAS XU system on Ikhana and a Thales 1030/1090 MHz
ground receiver, and compared against differential GPS data
as truth source for analysis. Representative segments of ADSB data were examined as to their persistence and consistency,
accuracy, and actual quality versus quality metrics (NACp and
NACv) asserted by the ADS-B system. In the limited sample
used in the analysis, the provision of position and velocity data
by ADS-B were persistent, gaps in surveillance due to noise or
between message cycles were sufficiently short to permit track
continuity through the gaps, significant errors were rare and
easily detected, and data accuracies were within the limits asserted in quality indicators. Surveillance quality was successful in meeting both ADS-B and TCAS standards. ADS-B should
continue to be considered a candidate for passive surveillance
for ACAS XU.

HONEYWELL STM USING SENSOR FUSION
The Honeywell tracking system (HTS) is a multisensor fusion
system that can simultaneously track up to 30 intruder aircraft
in real-time [6]. The HTS, based on a unique application of a
multiple hypothesis testing approach, fuses measurements and
statistical information from both cooperative and noncooperative surveillance sensors in one framework to track intruder aircraft in three dimensions. ADS-B and TCAS Mode-S provide a
globally unique ICAO ID for each intruder, but TCAS Mode C
and Radar provide sensor specific IDs for each intruder or no
IDs. Some of the largest challenges in sensor fusion research
56

are to correlate measurements from cooperative sensors with
noncooperative sensors and manage the IDs that accompany
the sensor measurements as intruders enter and leave the sensor's field of view. Each surveillance sensor also has unique
performance characteristics including measurement statistics,
accuracy, probability of detection/false alarms, field of view,
degradation based on operating environment, and update rate,
further adding to the challenge of sensor fusion. The prototype
HTS tested in this flight temporarily filtered out air traffic beyond a 20 nmi range to prevent any lags due to computational
calculation burden on the hardware.
The HTS is divided into three major subfunctions: data association, track manager, and tracking filters. Data association
assigns a cooperative or noncooperative sensor measurement
to an active intruder aircraft track. The track manager operates
together with the data association to initiate and maintain intruder tracks, coordinate ICAO IDs and sensor specific IDs for
active tracks, delete tracks that leave a sensor's field of view,
and reject false tracks. The tracking filter fuses assigned cooperative and noncooperative sensor measurements to estimate
the intruder trajectory statistics relative to ownship. These three
subfunctions operate together to estimate single intruder tracks
(relative position, geodetic position, relative vertical/horizontal
velocity, acceleration, azimuth, range, range rate and elevation
angle) from multiple sensor inputs and pass it to the ACAS XU
TRM and SAA display to determine whether any SS or CA maneuvers are necessary. These features of the multisensor fusion
system enable the UAS to safely operate in various mission
profiles and in mixed equipage situations across nonsegregated
airspace.
Figures 4 and 5 illustrate the incoming sensor measurements
and tracking performance of the HTS onboard Ikhana, respectively.
The two figures show the North-East-Down position of intruders
relative to Ikhana, as it was initially flying due west and performed
a horizontal maneuver to head due east between t = 200 to t = 285
at constant altitude. HTS was tracking six distinct intruder targets
simultaneously for this particular slice of data, but will focus on
one intruder (ICAO ID 11264663) to demonstrate its performance.
Figure 4 shows inputs from ADS-B, TCAS (Mode S), and radar
for ICAO 11264663. TCAS and ADS-B/radar reports routinely
come in at 0.2 Hz and 1 Hz, respectively. Radar measurements of
separate noncooperative intruders (radar IDs 72 and 73) are also
included to demonstrate the HTS's capability in differentiating in-

IEEE A&E SYSTEMS MAGAZINE

SEPTEMBER 2016

Table of Contents for the Digital Edition of Aerospace and Electronic Systems Magazine September 2016