The AWACS Surveillance Tracking System
The first demonstration application is an AWACS surveillance tracker, implemented jointly by the MITRE Corporation and the Open Group [Clark et al. 99].
AWACS and the Surveillance Tracking Problem
“The E-3 Sentry is an airborne warning and control system (AWACS) aircraft that provides all-weather surveillance, command, control and communications needed by commanders of U.S., NATO and other allied air defense forces. The E-3 Sentry is a modified Boeing 707/320 commercial airframe with a rotating radar dome. The dome is 30 feet (9.1 meters) in diameter, six feet (1.8 meters) thick, and is held 14 feet (4.2 meters) above the fuselage by two struts. It contains a radar subsystem that permits surveillance from the Earth’s surface up into the stratosphere, over land or water. The radar has a range of more than 250 miles (375.5 kilometers) for low-flying targets and farther for aerospace vehicles flying at medium to high altitudes. The radar combined with an identification friend or foe subsystem can look down to detect, identify and track enemy and friendly low-flying aircraft by eliminating ground clutter returns that confuse other radar systems. Major subsystems in the E-3 include navigation, communications and computers. Consoles display computer-processed data in graphic and tabular format on video screens. Console operators perform surveillance, identification, weapons control, battle management and communications functions.” [USAF Fact Sheet] The E-3 is illustrated in Figure 1. Some of the most recent AWACS aircraft are Boeing 767′s.
There are a number of different missions that an AWACS can perform. For example, it can perform air-traffic control, monitor and manage logistics such as refueling, or carry out general surveillance. The project discussed here needed to select a single mission as a case study, to fit within the project’s schedule and staffing constraints. Because of its general utility and intuitive simplicity, a surveillance mission was chosen.
When flying a surveillance mission, the AWACS operators attempt to monitor all of the airborne objects in a potentially large region. Once an object has been identified, the tracker should attempt to continue to follow its progress. The more closely the tracker’s estimate of an object’s position and heading agree with reality, the better. Moreover, once the tracker has identified a track, it should not “drop†it. A track is dropped if it is not updated for a number of input sensor cycles. This is inevitable if no new sensor input is received for the track. Often (the cases on which this project focused), sensor input is received, but is not processed – or is not processed correctly.
The multiple-target-tracking problem consists of detecting objects and following their movements based on periodic sensor data. The number of sensor reports can vary from radar sweep to radar sweep. Some reports are due to noise and clutter instead of being observations of aircraft. When new sensor reports arrive, a tracker associates the information contained in the sensor reports with the current estimated states of the tracks in its track file to produce updated track records that represent the tracking system’s estimate of the state of the airspace covered by its sensors [].
A typical surveillance tracking application is comprised of several stages. First it applies a divide-and-conquer technique consisting of a pair of algorithms called gating and clustering to split the association problem into smaller, mutually exclusive, subproblems, called clusters. Then data association solves each of these subproblems individually by correlating sensor reports in a cluster with tracks in that cluster. The final stages of a tracker – prediction and smoothing – are used to predict the next position, velocity and other parameters for each tracked object based upon its track history and the results of data association.
Data association is the heart and the bottleneck in tracking applications. There are many types of data association algorithms with varying levels of tracking performance and computational complexity. While the more sophisticated and computationally demanding data association algorithms tend to do better in more demanding multiple-target-tracking situations (i.e., large numbers of closely spaced tracks with similar motion in the presence of noise and clutter), their potential computational demands make them inappropriate for general use in the real-time tracking environment. They are especially problematic in precisely the situations in which they are most needed (i.e., areas with high track density), since the long computation time may lead to a timing fault and a tracking failure.
The trackers of interest are track-while-scan trackers, meaning that the tracker is getting sensor reports on successive scans of the airspace and correlates them across scans in both time and space dimensions to produce tracks. This correlation takes into account velocity, acceleration and other parameters of the objects being tracked. Consequently, if the reports corresponding to a given track are not processed during each sweep of a sensor, the position of the track becomes less certain and eventually the track will be dropped. A dropped track can be rediscovered; but this is a relatively costly operation and there is no assurance that any individual track will be reacquired.
The AWACS tracking system has a specified maximum track-processing capacity. The tracker processes data in first-come-first-served order, and thus fails to process the later data received from a particular sensor once its track-processing capacity is exceeded. Since sensor reports generally come in the same order from one sweep to the next, it is likely that, under overload, sensor reports from a specific region will not be processed for several consecutive sweeps. Without intervention, this overload behavior can result in entire regions of the operator displays that do not get updated – meaning that sectors of the sky “go blank,†as depicted in the Figure 2 cartoon.
Overload is a potentially serious problem because there is no inherent correlation between important regions of the sky and the arrival order of sensor reports. Currently, operators have knowledge-intensive manual work-arounds for certain overload situations. Higher performance computing only helps somewhat, because the potential workload and the need for ever-better tracking algorithms will always exceed the computing capacity.