Tactical Radar System Adds Teeth To Perimeter Security Missions

March 2000
By Henry Kenyon

Small, easy-to-use device offers alternative to anti-personnel mines by placing humans in target identification decision making.

A manportable sensor capable of detecting troops and vehicles up to 100 meters away offers commanders a variety of choices for defensive and surveillance operations. Consisting of a microwave Doppler radar unit and a passive infrared detector, the activated device transmits a message to a sentry who is equipped with a pager display that indicates the type of target and the direction it is moving.

Experts predict that the first decade of the new century most likely will see a continuing expansion of peacekeeping duties and operations other than war by the United States and its allies. Under these mixed conditions, maintaining security zones between combatants and providing a degree of early warning for forward troops on humanitarian missions becomes paramount. Current methods of area denial such as ground-based seismic sensors or anti-personnel mines are inefficient or politically unpopular. The move away from anti-personnel mines is an important issue for North Atlantic Treaty Organization countries because of recent international accords to ban the use of these weapons. New methods to detect intruders and deny opposing forces access to areas would allow commanders more operational flexibility.

A manufacturer of surveillance devices, Arkonia Systems Limited, Bordon, Hampshire, England, was asked by the British government to develop a remote ground sensor system for the British Army. According to Arkonia's managing director, Leonard Robinson, at the time his firm only made a passive infrared military sensor. After conducting an in-house study, company engineers concluded that existing sensors were not good enough to meet government requirements that included classifying a person or a vehicle with 90 percent accuracy.

The firm's engineers investigated a number of systems such as seismic, magnetic and audio detectors. They concluded that any item that relies on pressure media such as air- or ground-based vibrations was not adequate to accurately classify a target, Robinson observes. Likewise, while visual systems are the best for target identification, the use of cameras was eliminated as a possibility because of the requirement for operability in darkness and other poor visibility conditions. Visual recognition devices were also ruled out because the detector had to be manportable and easy to set up and operate.

Arkonia decided to pursue a radar-based method. The firm works with radar in other products and services it provides for the British government, so considerable expertise was available to research this area, Robinson shares. The Hornet sensor system, which was the product of this work, became available in mid-1999.

A small, self-contained and easily portable device, the Hornet consists of a passive infrared (PIR) detector mounted on top of a microwave Doppler radar module. The weatherproof, ruggedized body contains all major components, and 12 AA batteries provide power. The entire system, including the detector, tripod, bag and ancillary equipment, weighs less than 5 kilograms and fits into a soldier's backpack.

The PIR has a range of 100 meters. Any heat source entering the detector arc causes it to activate the radar, which emits a 3-second burst of signals. Return data is then analyzed against the Hornet's built-in classification library, and the information is transmitted to a soldier equipped with a pager. The signal tells the sentry the target type and bearing.

Arkonia officials claim the device is 90 percent accurate, but Robinson says the actual accuracy level is about 99.4 percent. These levels are attained with the current version of the Hornet, which operates on a single channel. The next developmental step is to replace the Doppler radar with a quadrature system, which is less susceptible to background clutter.

The Hornet also features an advanced infrared sensor. Standard PIR detectors normally operate by creating a standing voltage. Any heat source that reaches the detector will cause a change in that voltage, triggering a target registration process. Robinson notes that this can lead to a variety of problems. For example, the sun emerging from behind a cloud could register a false alarm. By comparison, the Hornet's PIR is a digital device consisting of right and left detectors that create a data stream. Based on the time constants of the incoming information, the sensor can ignore phenomenon such as bright sunlight or rain. The sensor has given no false alarms to date, Robinson claims.

If the Hornet detects multiple heat sources and it cannot decide whether they are personnel or a vehicle, it will register them as a target. This has occurred twice in hundreds of radar passes, he notes. In this scenario, the device has a logic output that can trigger an optional small video camera mounted underneath the sensor to send an image back to a command post for verification of the target.

For area denial, the detector covers a triangle 100 meters long by 70 meters wide. However, this front can be expanded up to 200 meters by simply turning the device on its side, which enlarges the beam width, Robinson says. The power output is also increased from 10 to 100 milliwatts to provide this additional coverage. "All we have to do is put it on its side because, with radar, the wider the horn [the coverage arc], the narrower the beam. By turning this horn sideways, we use the narrow part of the horn, which means a wider beam," he notes.

The next step for the Hornet is the inclusion of quadrature radar. This radar operates on two channels--one positive and one negative. When combined, they cancel each other out. This enables the system to ignore background clutter caused by certain kinds of natural movement--for example, trees bending in the wind or waves breaking on a shoreline--because they are only moving in two vectors. However, a person or a vehicle would create a third vector as it entered the detection area. This cannot be canceled by either channel and would register as a target. The Hornet's processor would then transfer this frequency data to determine whether the signal was a human, a vehicle or a helicopter. This quadrature application has proven successful in trials, Robinson adds.

The logic power for the Hornet comes from an Arkonia-designed real-time fast Fourier processor that allows the system to accurately identify a soldier crawling on the ground at 100 meters. In order to classify targets such as people, up to 400 radar passes were made of people walking and crawling in all directions. From this data, a mathematical model was created for human movement. The same was done for a variety of vehicles and helicopters. All of these models are stored in the Hornet's classification library. If something is detected that does not meet the device's criteria--a dog for example--it will be identified as a target. Robinson adds that these also would be considered a nuisance alarm unless the system was configured to ignore them.

An example of this kind of nuisance alarm involves kangaroos. The Australian government is interested in using the Hornet in an area where wild kangaroos are present. So that the animals do not set off nuisance alarms, a pattern must be created for the system to identify and classify. Doing this in a zoo would not capture the type of movement found in the wild, so Arkonia went to Australia to collect data on the way feral marsupials move. Because the device has been available for slightly less than a year, the pattern library still needs substantial additional data. Robinson adds that in the future, Arkonia might do the same type of research with dogs.

The Hornet currently is in use with the Macedonian army and is being assessed by the British Army and the U.S. Federal Bureau of Investigation. Though it has much promise, the current version of the device does not meet the British Army's needs for area denial, according to Maj. Jeremy Stadward, British Army, staff officer grade 2 (weapons) surveillance, U.K. Ministry of Defence's Surveillance Target Acquisition Night Observation and Countersurveillance (STANOC) Centre. Instead, the army is waiting for the company's next generation of detectors that will feature quadrature radar because it is much better at detecting people, he says.

Noting that his opinions do not reflect those of the STANOC Centre, the major adds that the British government was seeking replacements for its existing sensors and tapped Arkonia for its expertise in the surveillance field. Pressure to find new methods for area denial was also growing. "What a lot of people are looking for is a replacement for anti-personnel mines," he says.

Beyond alerting sentries, the Hornet can be used to help activate a command-detonated device placed in the security zone. Unlike anti-personnel mines, which are indiscriminate, the Hornet would put a person into the verification loop before a weapon is activated. This is where the area denial concept would really have meaning, Maj. Stadward observes.

The Hornet's ease of use also presents advantages over seismic systems such as geophones, which detect vehicles or personnel by the ground vibrations they make. Setting up a perimeter with such a system involves substantial work and time because the devices must be dug into the ground. According to Robinson, it takes five ground sensors to match the coverage of one Hornet, and they must be continuously monitored by an operator. Geophones also produce false readings and have a classification accuracy of only 40 percent under favorable conditions. Robinson notes that environmental conditions such as tree root movement or soil type adversely affect readings. To compensate for background noise, operators often turn geophone sensitivity down, which means that intruders are not always detected in the security zone.

By comparison, the Hornet requires little or no training to operate. A soldier assembles the equipment--which takes about 30 seconds, Robinson estimates--points it in the desired direction, and plugs it in. The device is turned on when the PIR is plugged into the radar sensor. Instead of providing a sentry with a pager, the Hornet can also be connected via a land line to a computer, or it can broadcast by radio to a central location.

Another enhancement to the Hornet that Robinson sees for the near future is what he calls sensor scouts. Currently, the device relies on the PIR to activate it. This saves power and prevents detection. The sensor scouts will be deployed 100 to 150 meters from the radar sensor. These devices detect the magnetic fields of any person or vehicle passing within 20 meters. A signal is sent back to the Hornet, which then switches on, classifies the target, and sends the information to a sentry or command center. Robinson adds that these devices can be used with or without a PIR, or they can be placed to cover parts of an arc. The firm has the ability to build the sensor scouts now; it is simply a matter of filling out existing orders before Arkonia has the capacity to implement their production, he says.

Besides area denial, the system has other applications such as border surveillance or headquarters protection. Robinson notes that one application being provided for a foreign customer is classification. For example, in an anti-narcotics operation, a Hornet may be concealed near a small airstrip. Because the device can be programmed to ignore certain types of movement such as people or automobiles, it will only transmit an alarm when it detects a small aircraft.

The sensor recently assisted in crime fighting, Robinson says. A local firm was losing equipment to theft. The only way this could be easily accomplished was if the goods were loaded into a van at night. A Hornet system configured to detect a van was put on the side of the building. After several months, a van arrived one night. The device recognized it, set off an alarm and the thieves were apprehended.