Radar development and tracking technology go parallel

For the past couple of years, relays of quad bikes have been driving around Stonehenge in Wiltshire towing trailers loaded with electronics in an attempt to see what our ancestors buried there thousands of years ago – but without having to send in teams of diggers.

The Stonehenge Hidden Landscapes project, begun in 2010, has already uncovered evidence that has reshaped our understanding of how the enigmatic circular stone structures were built and used, with repeated surveys being run to try to find out more. Up to a couple of years ago, various surveys had mapped out features over an area of 4km2 using traditional digs. By deploying an array of electronic instruments, from ground-penetrating radar (GPR) to laser scanners and magnetometers, the project aims to survey almost the whole World Heritage Site – some 20km2 in all.

Non-contact techniques for surveying the underground are not restricted to archaeology. The Engineering and Physical Sciences Research Council (EPSRC)’s Mapping the Underworld research project draws to a close at the end of July and seeks to provide technology that can cut the amount of time and money spent digging up roads simply to work out where subsurface cables and pipes lie – but the ground does not give up its secrets readily to techniques such as ground-penetrating radar.

“You have to adapt to different ground conditions,” says Dr Steve Pennock of the University of Bath and one of the principal researchers on Mapping the Underworld. “Some types of ground are lossier than others and have different frequency responses.”

A sweep across a wide frequency range could reveal a lot, but in the 21st century there are acute restrictions on what you can do. The instruments suffer from the various powerful transmitters that work in the range needed by GPR, Pennock reports. “GPRs typically have to work at less than 1GHz,” says Pennock, “but in that range you can have powerful TV, cellular telephony, and radio signals that will interfere with the radar. You have mobile phones, for example, around 900MHz but it is narrowband, so you can clip that out or it will misguide the radar. You also have the TV signal to deal with, and that varies from area to area.”

By making the radar adapt to the changing radio environment, it can avoid trying to use the frequencies of the strongest interferers in the vicinity. Although there will be gaps in the response, it means that with one sweep an instrument could find out a lot about what is underneath it. There are number of ways to achieve that, but it is also an area where technology and theory have yet to fully converge.

Modulated frequencies

The technique that potentially offers the best performance – derived from the transmission techniques used by conventional Wi-Fi – is orthogonal frequency-division multiplexing (OFDM). This makes it possible for digital signal processing (DSP) to distribute the transmission across a set of frequencies, and simply knock out those affected by interference.

The problem for commercial radar systems is that, although the DSP horsepower is available, other components, such as the analogue-to-digital and digital-to-analogue converters, are not. “You need a 1GHz ADC and DAC, but there is no commercial drive for that,” says Pennock. “We did an estimate based on what is available today and the rate of development, and I would have retired before they became available.”

Other techniques are available, such as frequency-modulated continuous wave (FMCW), similar conceptually to the technique used in commercial perimeter-security systems such as those made by surveillance solutions firm Blighter. The signal sweeps up or down from one frequency to another. They are less flexible as you have to manage the notches for cellular and TV in a more complex circuit design but they are achievable with today’s front-end components.

Another option is to bury the transmitter itself – putting it into a ‘mole’ that moves through an underground pipe. Interference could cause problems here as well. “With commercial ground-penetrating radar, the transmitter and receiver are almost in the same place, which does tend to satisfy the regulatory requirements because the intention is to radiate into the ground,” explains Pennock.

Transmission from the ‘mole’ is potentially more problematic, because the waves will propagate out from the surface and potentially interfere with other systems – something that the regulatory agencies frown upon. Yet the notion of separating radar transmitter from receiver is gaining interest, as military and commercial radar users try to eke more out of the limited frequency spectrum they themselves have to deal with.

The popular image of the radar is the rotating antenna that sits on top of a tower at an airport or on a ship. But this traditional design is arguably reaching the end of its development potential. The steady rotation, even for an antenna that rotates once a second, is too slow to pick up a supersonic aircraft or missile – by the time the radar has completed a single rotation the object is potentially more than a mile closer.

The generations of radar systems developed over the past couple of decades, such as the Artisan radar fitted to HMS Iron Duke by BAE Systems last March, and which will go onto a variety of naval frigates, combine this mechanical rotation with electronic scanning. This uses constructive and destructive interference from an array of transmitters to steer beams electronically towards likely targets. The latest generations of systems build-in jamming resistance by allowing the radar to produce a much broader spread of frequencies and so avoid concentrated interference. 

Techniques being explored by researchers such as Professor Hugh Griffiths of University College, London (winner of the 2012 IET AF Harvey Engineering Research Prize) could result in radar systems that are extremely difficult to jam because the enemy does not know where half of the radar actually is. Contemporary radars bounce electromagnetic pulses off targets and pick up the reflections with antennas at the same site. But there is no need to put the transmitter and receiver in the same place. You don’t even have to use your own transmitter, you can piggyback off FM radio or TV transmissions.

A radar that went into service in the late 1990s on Manastash Ridge in Washington State (US) uses FM radio transmissions to help map disturbances in the upper atmosphere. For more mainstream uses, the benefits of moving the transmitter and receiver apart have not, so far, been worth the additional complexity of design; but things are changing.

Origins of radar innovation

The idea for a split radar design dates back a lot further than the 1990s. The first operational system was put together during the Second World War. But for a mistake made by German scientists before the first shots were fired, it could have proved to be a thorn in the side of the Allies as they prepared the Normandy invasion.

The Germans sent an airship to fly over the North Sea one month before the outbreak of war to probe British defences. Aware that the British had some form of radar technology, German forces wanted to work out how widespread it was. They assumed that strong megahertz radio signals the airship picked up were too low in frequency to be radar signals. They were wrong: they were the emissions from the Chain Home radar system that helped the British pinpoint incoming aircraft on bombing raids during the Battle of Britain in 1940.

“If they had associated it with radar, things might have been quite different,” UCL’s Griffiths says.

A year later, the Germans changed their minds about the signals, but it would be several years before they put together a cunning plan to use Chain Home’s own signals to detect Allied aircraft flying towards the continent. The Klein Heidelberg antennas scattered along the English Channel and North Sea coasts picked up echoes from the Chain Home network of transmitters and relayed them to trained operators. According to Griffiths, who has used declassified records to piece together the history of Klein Heidelberg, the signals received by this type of bistatic radar needed interpretation.

Although, he says, “German radar operators then tended to be people who weren’t fit for other duties and so weren’t the brightest”, the additional complexity of deciphering the signals from this kind of radar did not pose a problem to pinpointing aircraft.

Klein Heidelberg’s impact on the war was limited for reasons that remain unclear, although with deployment only starting in the late spring of 1944, it could simply have been a case of too little, too late. German commanders seemed to have had little faith in the system: one of the receivers was apparently dismantled just a month after it went into service.

Griffiths argues that the technique can provide the impetus for a new generation of radar systems by doing things conventional radar systems cannot. For example, a stealth bomber is designed to scatter radio pulses rather than bounce them back towards the transmitter. Receivers placed far away from the transmitter could detect that scattered energy.

“There is the possibility that bistatic radar may be better at finding stealthy targets,” says Griffiths.

One further advantage is that, by removing the power-hungry transmitter from the receiver, it is possible to make the sensor-end of the radar more mobile. “We are very interested in using UAVs now and bistatic operation is very compatible with all that,” says Griffiths.

There are further degrees of freedom available to radar designers if they move away from traditional architectures. Multistatic and multiple-input, multiple-output (MIMO) designs offer new approaches that effectively build a mesh radar out of individual systems. One of the problems with multistatic architectures is that you need good information not just on the position of transmitters but on the timing and type of pulses they send. It’s hard to know where an echo should be without that data. Today’s UAVs (such as General Atomics’ Predator) need self-contained radar subsystems, but future designs could save precious weight and fuel by moving to multistatic architectures in which only the receiver needs to be airborne. Access to satellite-navigation systems provides not only good positional information but timing – the Manastash Ridge radar uses the Global Positioning System (GPS) to synchronise its multiple receivers.

By using more diverse signals, it is possible to home in on targets designed to evade traditional methods and avoid jamming signals. One issue that today’s radar systems have is that they are constrained in the frequencies they can use. One option might be to build radar systems that sniff the airwaves to dynamically adjust to use pieces of unused spectrum. Even regular TV and radio signals could be adjusted to make them compatible with radar, Griffiths says.

“Everyone wants more bandwidth, but it’s increasingly difficult to fit everyone in. So there is a lot of interest in techniques that might use broadcast signals as radar sources. You could even think of designing your broadcast waveforms so that they do not just do their normal job, but are also made more useful for radar,” says Griffiths.

“The key is identifying the right applications. Air-traffic control looks to be an interesting one. You could use broadcast signals for that. In densely populated areas, suitable radar exists; but in some areas there is nothing. Broadcast signals are strong and their transmitters are favourably located to give high coverage. In Africa, for example, being able to use broadcast sources for radar would give you something you don’t have yet.”

The technique scales down too; burglar-alarm systems of the future may piggyback off wireless network signals to determine if an intruder is moving around inside or outside the premises. Splitting the radar in two is set to put it on its next wave of development.

Integration of sensors

Sensor fusion is providing a way for the radar to move into everyday use. Cars are now beginning to sport active radar systems; but, as David Price, CTO of vehicle-technology specialist Pi Innovo, points out, radar has its limitations. Rain and moisture in the atmosphere reduce the penetration of the 60GHz signals – much higher than those used by many long-distance military systems. “Radar, optical, and infrared sensors are being integrated. You will get multiple systems on an individual vehicle because some fare better than others in different weather conditions,” Price explains.

At the Embedded World show in Nuremburg in February 2013, programmable logic solutions firm Altera Corporation demonstrated a combined radar and optical-camera system that uses its programmable-logic devices to combine the inputs dynamically.

Built and programmed by components specialist Data Device Corporation, the system tracks objects in the path of the vehicle, working out whether the distance from the car to each object is rising or falling. That information can flag up potential hazards to the driver or be used by other computers in the car to handle active cruise control or even to apply the brakes automatically if the machine thinks a collision is imminent.

Multistatic radar and sensor fusion may even be adapted to medical diagnosis. Researchers at the University of Bristol’s Centre for Communications Research have looked at using techniques used in landmine detection – some of which combine ground-penetrating radar with multistatic analysis. Malignant tumours tend to scatter microwave-frequency signals more strongly than healthy cells. Meanwhile, a team at Canada’s McGill University has proposed combining microwave radar with acoustic techniques to listen for the difference in high-frequency sound transmission through tumours and healthy tissue.