design considerations for adaptive active phased-array 'multifunction' radars

Design considerations for adaptive active phased-

array ‘multifunction’ radars

by John Holloway

Adaptive active phased-array radars are seen as the vehicle to address the current requirements for a true ‘multifunction’ radar system. Their ability to adapt to the environment and schedule their tasks in real time allows them to operate with

performance levels well above those that can be achieved from conventional radar designs in addressing the current and future threats. Their ability to make effective use of all the available RF power and to minimise RF losses also makes them a good candidate for future very long range radars. This paper addresses the design of such

radars in terms of the system and i t s component parts and the operational requirements that drive the design. The paper also considers some futures uses

outside the military field.

Abbreviations

AAPAR = Adaptive Active Phased-Array Radar ASDE = Airport Surface Detection Equipment ATC = Air Traffic Control ATM = Air Traffic Management BIT = Built-In Test CFAR = Constant False Alarm Rate ECM = Electronic Counter Measures EIRP = Effective Isotropically Radiated Power ESM = Electronic Surveillance Measures IFF = Identification: Friend or Foe LNA = Low Noise Amplifier LRU = Line Replaceable Unit MFR = Multi Function Radar MSSR = Monopulse SSR NCTI = Non-Co-operative Target Identification PRI = Pulse Repetition Interval RF = Radio Frequency SSR = Secondary Surveillance Radar

1 introduction

Over the years radar systems have been developed to take account of changes in requirements caused by:

(U) increases in the number of wanted and unwanted

(b) reductions in target size either due to physical size targets

reduction or due to the adoption of stealth measures

(c) the need to detect wanted targets in ever more severe levels of clutter and at longer ranges

(d) the need to adapt to a greater number of and more sophisticated types of electronic countermeasures (ECM, jamming).

Radar designers addressed these needs by either design- ing radars to fulfil a specific role, or by providing user- selectable roles within a single radar. This process has culminated in the fully adaptive radar, which can automatically react to the operational environment to optimise performance.

The ‘detection problem’ is illustrated in Fig. 1. The radar has to ‘see’ and identlfy the target echo amidst a host of natural and man-made echoes, interference and jamming.

Conventional radars fall into two broad categories independent of what functions they perform. The first category has fixed antenna patterns produced by either reflector or passive array antennas, with centralised transmitters. The beam being fixed, scanning can only be achieved by physically moving the antenna. Typically a surveillance radar will produce a fan-shaped beam with a fured elevation illumination profile, the azimuth scanning being achieved by rotating the antenna. A tracking radar will have a pencil beam that is used to track targets by the use of a mechanical tracking mount. Because of the limitations imposed on such radars by their design, such radars are ‘single-function radars’. The basic layout of a single-function radar is given in Fig. 2.

The second category of radars is the passive phased

ELECTRONICS & COMMUNICATION ENGINEERING JOURNAL DECEMBER 2001 277

the operational requirements that drive the design.

2 Background

Tu rget size Radar echoing areas have

become smaller through physical size reductions, modern materials and the introduction of stealth techniques. In the case of piloted aircraft, sizes have

Fig. 1 The detection problem

array. These incorporate electronic beam scanning or beam shaping by the use of phase shifters, switching elements or frequency scanning methods. These features enable the radar designer to implement more complex systems having the capability to carry out more than one radar function, i.e. ‘multifunction radars’. Generally, however, the functions of the radar are preprogrammed and not adaptable as the radar environment or the threat changes. These systems traditionally use centralised transmitters and their limited amount of ‘multifunction- ality’ is bought at the price of considerable beamforming losses. The layout of a conventional passive phased array radar is given in Fig. 3.

In order to improve the multifunction capability over that of a conventional phased array, in many cases the adaptive active phased-array radar (AAF’AR) is the only practical solution. In the MAR, transmitter/receiver modules are mounted at the antenna face and adaptive beamforming and radar management and control techniques are employed. These techniques provide the current state-of-the-art solution and result in a truly ‘multifunction’ capability.

This paper discusses the design of such M A R S and

fallen from typically 10 m2 to 1 m2, and are claimed to be

<< 0.1 m2 for stealth aircraft; in the case of missiles, areas vary from 1 m2 to 0.01 mz or less. This has lead to a need to increase the detection performance of radars.

In parallel with this reduction in target size the effectiveness of weapon delivery systems has improved substantially. The range at which munitions can be released has increased. This, compounded by the increased speed and lethality of modern weapons, has led to a commensurate increase in the range at which targets need to be detected.

This has resulted in the current situation, where in a lot of cases it has not proven possible to adopt the traditional remedy of increasing the antenna size and transmitter power to improve target detection, due to the limiting nature of the clutter. Other methods need to be adopted to detect modern small fast targets in clutter, with an acceptable level of false alarms.

The size, complexity and prime power requirements of a conventional radar able to detect modern targets generally limits the radar to a static installation. In the case of naval platforms the ship has to be sufficiently large to accommodate the radar. This need for large installations conflicts with the current requirement for

radars that can be deployed rapidly away from base.

data to system

displav 1

target

1 continuous rotation

Fig. 2 Conventional radar design

Environmental considerations Along with changes in target

characteristics there has also been a major change in the radar electromagnetic (EM) environment. This consists of natural elements-land, sea and weather clutter etc.-and man-made elements, such as background interference, mutual interference from other systems and ECM. It is in this ECM area that the major changes have occurred. In the future, however, a prolieration of other radio-based services is likely to raise the level of background man-made interference. This is particularly true for systems operating in conflict

278 ELECTRONICS & COMMUNICATION ENGINEERING JOURNAL DECEMBER 2001

areas bordering neutral states where there is no control over civilian RF transmission.

The effects of natural clutter on radar performance are well known and standard techniques, of varying effectiveness, have been developed for conventional radars to deal with these effects. These techniques, however, are generally established at the radar desigdintegration phase and are generally imple- mented by a number of firmware or software operator-controlled options. These generalised sett- ings, in the case of static radars, can provide a good capability against fixed clutter where the radar parameters are optimised on site. However for mobile radars operating at variable locations and for fured radars operating in varying weather clutter conditions the results are less than optimum.

Over the years the design of ECM systems has become much more effective and radars have had to become more sophisticated in order to counter them. As in the case of natural clutter the methods used to defeat ECM have usually been provided as a series of predetermined functions. It has not proved possible to adapt the radar parameters quickly to cope with the changing ECM environment.

In the short-term, conventional radar parameters cannot easily be adapted as the ECM threat changes throughout a mission. In the longer term the radar design needs to be constantly updated to cope with a change of types and numbers of ECM equipments.

high-power target

1 display I

radar plots

variable phase shiners

arrav I ;example oi I 9 radiators, but I usually 1000+) I

_ - - - _ _ 4 . computer I- operator/system beam control requirements

Fig. 3 Conventional phased array

display 1 !

solid-state variable Txand phase duplex, SI hiftei

I I

ar

I

I I

/ planar

target

array

I I I

I I operator/system

requirements L - - _ _ _ - - _ _ _ - - - _

control

radar plots

Adaptive active phased-array radar Fig. 4 Active phased-array radar

Active arrays: A major reason for the large size and power requirements of a conventional phased-array radar is the need to overcome the loss in the RF signal between the bulk transmitter and the antenna, and between the antenna and the receiver. Losses typically can be 7 dB (2- way) and in some complex designs can reach as much as 10 dB. Typically 95% of the prime power and 80% of the effective transmitter power is lost, with only 20% being used for detection. This may be acceptable for static systems running from a dedicated mains electrical supply but is not acceptable for mobile systems or those running off standby generators.

Combining in space the power of many low-power radiating modules, mounted on an antenna face as in an

MAR, ensures that the power is radiated directly into space with the minimum of loss. If the same modules are used for reception with a low-noise amplifier (LNA) stage close to the array face, then similar reductions in receive losses are obtained. This gives active arrays a major benefit in pure detection performance. Prime power requirements are also greatly reduced, allowing the use of smaller generators in mobile systems and reducing power consumption costs in static systems. A typical layout for an active phased array is given in Fig. 4.

Adafitive radarfeatures: The use of active modules provides the ability to control the radiation and receiver parameters of an active-array radar in real time, and to adapt these as the threat changes. Adaptive radar features


Fig. 5 Volume surveillance

are added to an active array to produce an MAR. Features that can be adapted include:

digital beamforming; 0 waveform generation and selection

beam management frequency selection task scheduling

0 tracking.

This has for the first time provided the system with the ability to optimise the parameters of a radar to meet natural and environmental effects as they occur. Human operators may be used to adapt the system, however generally not in real time. The system can be programmed to adapt automatically as the threat changes, using a set of rules-either preprogrammed or selected by an operator.

The increase in performance of an active array radar within the environment and its improved detection performance over conventional radars make the active- array radar highly versatile and flexible in operation. It is now possible to design a radar to react to changes in the threat scenarios and to adapt its own parameters to optimise performance.

3 Operational requirements

Radar roles The roles of the radar sensors in a typical air defence

system need to be specifled in order to define what the AAPAR is required to do. In a conventional current design, a combination of a surveillance radar and one or more tracking radars is normal. The r81e of the defence system, i.e. its purpose or mission, will dictate the key attributes required of its sensors. A short-range point defence system will have significantly different characteristics to those of a long-range system designed for, say, ballistic missile defence.

This paper does not consider detailed parameters for the differing types of radars, rather it discusses the generic functions required in terms of defining a multifunction radar of an AAPAR type.

A radar sensor as part of an air defence system may be required to perform a number of functions in order to generate and maintain target data and to assist in engagement of targets. The principal functions are:

volume surveillance target detection and confirmation target tracking target identification by both co- operative and non co-operative methods target trajectory or impact point calculation

0 tracking of ECM emissions kill assessment

missile or other communications.

Volume surveillance The MAR can provide a number of operating modes

to tailor the surveillance volumes to the system or mission requirements. Energy usage is optimised and the probability of target detection is maximised by management of the radar waveforms and beams. Volume surveillance can be managed in order to cope with varying threats-lower priority surveillance tasks (long-range search perhaps) can be traded for higher priority tasks such as short-range surveillance or target tracking as the threat scenario changes, see Fig. 5. In the limit the system could revert to a point defence mode, abandoning area defence considerations altogether.

Detection and confirmation A look-back beam using the position data derived

from the detection beam can immediately confirm each detection that is not associated with a target already in the current track file, thus significantly reducing the track confirmation delay.

Target tracking Separate tracking beams can be used to maintain target

positions and velocity data. The target parameters and priorities determine the update and dwell times. Targets with low manoeuvring capability or those that are classi- fied as friendly or neutral may be tracked using track- while-scanning techniques during normal surveillance.

Detailed tracking requirements are defined by consideration of the complete process of target interception. The architecture and capabilities of the interceptor element of the system and the range at which the final interception takes place largely drive the requirements of the tracking radar.

Target identt3cation Co-operative techniques use an IFF (identification:

friend or foe) interrogator system controlled by a radar. Depending on the role of the radar, interrogation of targets is performed only when demanded, or on a continuous ‘turn and burn’ basis. Selective interroga-

280 ELECTRONICS &I COMMUNICATION ENGINEERING JOURNAL DECEMBER 2001

tion is used to minimise emissions from the radar to reduce the probability of ESM (electronic sureveillance measures) intercepts and is nearly always used when Mode 4*, the secure IFF mode, is being used. (Turn and burn is used when the system is acting more in an air tt-a€fic control (ATC)/secondary surveillance radar (SSR) role, using Modes 3/A and C.) By confirming a target by looking immediately with the IFF mode, target confirmation times are reduced.

Non co-operative techniques extract additional data from radar returns by extracting features and comparing them

Fig. 6 ATM multifunction radar

with information held on threat databases. A correlation process is used that finds the best fit to the data. This method can provide good accuracy in recognising a target from a class of targets, or as a specific type of target.

The data and the models vary depending on the recognition technique being employed. However the collection of the target data generally requires an extended dwell. This is most effectively accomplished with an W A R , which can adjust the dwell time to optimise the data extraction.

Target trajectory calculation Calculation of an impact point is one input to the threat

assessment process, and the radar can assist by adapting to a mode that fits the trajectory to a complex curve-fitting law. This process is more effectively performed by the AAPAR since it can adapt its tracking priorities and parameters and form the data quickly to the required accuracy.

Intercepts of certain classes of target are only successful if accurate intercept points can be predicted. The accuracy demanded of the prediction may warrant long-range dwell times. Under these conditions the tracking process is optimised to the demands of the interceptor system, data rates and dwell time can be adapted to the requirements of each phase of the interception and many interceptions can be controlled simultaneously.

In other roles, the radar can use trajectory fitting to back-track to the launch point of the ballistic target for weapon location.

Changing the radar parameters in this way can only be done by an adaptive system. *IFF/SSR Modes 1, 2, 3 and 4 are used by the military and Modes A, C and S by civilian operators. In Mode A/Mode 3 the transponder replies by transmitting its code. In Mode C the transponder gives its altitude. Mode S (selectivity) is based on unambiguous identification of the aircraft by its 24-bit address and Mode 4 is encrypted.

Tracking of ECM emissions Receive-only beams can be formed with an active array,

giving all the normal receive processes without the need for transmitted W. Utilising these beams, sources of in- band radiation can be accurately tracked in two dimensions. The track data can be correlated with strobes from other sensors to enable the positions of jamming sources to be determined and tracked in conditions in which the presence of jamming may prohibit the formation of tracks.

Kill assessment It is possible to use a radar sensor to give some

information to the kill assessment process. The radar can be used in two ways. Firstly it can determine whether the trajectory or track vector has changed sufficiently to indicate that the threat has aborted its mission or been damaged sufficiently to lose control. Secondly the radar can form a high-resolution image of the target to determine if it has been fragmented.

Missile communications In a system where an interception is being performed

by a surface-to-air missile the multifunction radar is likely to be located in a position where it has good visibility of both the target and the outgoing missile. In this system the ground-to-missile communications link, used to control the missile in its various stages of flight, could be performed by the radar.

ATM (Air Traffic Management) multahnction radar So far, the use of the AAPAR has been considered for

defence applications. Generally it has been considered that only defence systems can warrant the complexity and the associated costs of the W A R . Recent work, for example the Federal Aviation Authority’s TASS (Terminal Area Surveillance System) programme, has considered the use of a multifunction radar to fulfil all the requirements of terminal area surveillance. It is possible to


the target scheduling compared to that required by a rotating system. The use of electronic scanning in the selective addressing process has the secondary benefit that it effectively smoothes out the interrogation rate peaks that occur in a conventional rotating system. These peaks result from air traffic bunching in sectors, particularly in approaches to airports. In such cases the majority of targets that require interrogation will lie in narrow approach and departure corridors. For a rotating system this leads to very high peaks in the interrogation rates required, placing a very high demand on the transmitter for a short period.

An AAPAR can operate with an effectively constant interrogation rate, resulting in a constant transmitter duty. This is particularly important

Fig. 7 AAPAR block diagram

conceive of an AAPAR system that carries out the following ATC roles:

terminal area surveillance medium-range/en-route surveillance general movement and tracking MSSR (Monopulse Secondary Surveillance Radar) Mode S data link ASDE (Airport Surface Detection Equipment) precision parallel runway monitoring weather surveillance: -precipitation; measurement and classification -wind sheer and microburst detection wake vortex detection.

when a Mode S data link is considered. Data link messages have to be scheduled along with surveillance interrogations; data link messages however do not arrive in a time frame related to the position of a rotating antenna. This means that, in the worst case, messages could be delayed by a period related to an antenna scan period.

Adding data link messages that are much longer than surveillance messages leads to a requirement for an instantaneous duty of ~ 7 0 % for a rotating system. Electronic scanning can be used to smooth out the interrogations in time, leading to a much lower duty of <lo%. Electronic scanning also results in a much more

The wide-band nature of the MAR architecture makes it possible to integrate the SSR into the primary radar.

In SSR systems Mode S is replacing the current Modes A and C. Mode S is a selective addressing system and the ability of an MAR to arbitrarily scan the beam simpliies

timely message transmission system, removing as it does the need to wait until the antenna rotates to the target bearing.

Currently one limiting factor has been the layout of the airport. Unlike defence radars, which can be sited in

an optimum position, airport radars have to fit the existing site layout. This generally limits the possible use of a multifunction radar due to shadowing and the relative positions of the runways, flight paths etc. However, with the rise in air traffic and the need

transmitter power for new green field airport sites (constant beamwidth) the possibility now exists for

siting a multifunction radar in an optimum position. Such a radar could be capable of true ‘gate to en-route’ operation, see Fig. 6.

4 AAPARdesign

System design To perform its multifunction

Fig. 8 Frequency compromise role the AAPAR is required to


carry out the following processes:

signal generation 0 transmit 0 receive

beamforming signal processing

0 tracking 0 data extraction 0 radar management 0 power and cooling.

These processes may appear similar to those in a conventional radar, however the detailed implementation in an M A R is fundamentally different and provides the flexibility required for the radar to perform the multifunction role.

A simplified block diagram for an MAR is shown in Fig. 7. In principle the M A R configuration in Fig. 7 is the same as the block diagram of a conventional radar. However the radical differences in beam management mean that the signal processing of an MAR is closer to that of a tracking radar than that of a surveillance radar. The other obvious difference is in the construction of the transmitter/antenna/receiver chain.

Performance drivers The design of an MAR is driven the same way as a

conventional radar by the types of targets it is required to detect and their ranges and properties. Because of the adaptive nature of the radar a much wider mix of target

0.75

0.50

0.25

0

-0.25

-0.50

-0.75

-1 .oo

types can be accommodated and the mix can be varied as the scenario changes. However the laws of physics still apply and the radar needs enough time and power to accomplish a detection. The design of the MAR can be optimised to make the best use of the time and power available such that maximum performance can be achieved in any given target mix. The system can also be programmed to prioritise roles and to 'turn off' functions as the target load increases in order to provide more time and power to the more critical functions.

The typical design drivers that have to be accommodated are:

stealth, i.e. very low radar cross-section targets 0 rapid reactiodupdates

highly manoeuvrable targets multiple targets very low sea-skimming targets intense jamming severe clutter

0 weight and prime power limitations 0 mobility and transportability.

Choice offiequency Operatingfi-equency: The choice of the radar frequency

-usually in the range 1-20 GHz for medium range weapon systems-has a major impact on the performance of the radar and its physical characteristics. The curves in Fig. 8 show the general trends. Clutter is a key performance limiter and tends to increase rapidly with radar frequency (e.g. rain return is a 4th power law with

-1.00 -0.75 -0.50 -0.25 0 0.25 0.50 0.75 1.00

Fig. 9 Contour plot showing main beam and grating lobes


Fig. 10 Adaptive beamformer

frequency) and consequently radar designers try to use as low a frequency as possible. The antenna aperture is chosen to provide the required beamwidth (determined by angular accuracy requirements) and is made as large as possible so as to give the maximum transmit EIRP (Effective Isotropically Radiated Power) and receive gain consistent with the largest practical physical size. Physical size and weight limitations-especially for naval or mobile/transportable systems-set the practical limits on how large the antenna can be.

In an active array it is the EIRP that needs to be considered because the directivity and total transmitter power are directly linked. The gain and the power radiated are a function of the number of antenna modules, which is directly related to antenna area and gain.

Size can be reduced by increasing the radar frequency, but this is at the expense of increased transmitter power and more difficult clutter suppression requirements.

The practical difficulties of cooling RF power modules and their inherent cost also increase nonlinearly with frequency.

Target size is tending to fall, in particular due to the use of stealth techniques. This requires even more transmitter power to achieve a signal return greater than the noise to ensure that the target can be detected. Given that, in practice, transmitter efficiency, and hence output power, tends to fall with increasing frequency and that

stealth techniques are less effective at lower frequencies, the operating frequency is therefore chosen as low as possible consistent with physical size constraints.

A simplified method for choosing the frequency is as follows:

(a) Decide on the beamwidth/aperture required based on a compromise between tracking, surveillance and clutter.

(b) Select the minimum number of elements to fill the aperture based on the beam scanning requirements.

(c) Select the lowest frequency band based on the constraints on the aperture size required for the number of elements.

(d) Select the lowest power module based on the required detection performance.

Operating bandwidth: The operating bandwidth and the number of operational frequencies is a function of the roles specsed for the radar. Potentially an AAPAR can have an overall bandwidth of up to 25% of the carrier frequency and can operate with pulses ranging from very short uncompressed pulses to long expanded pulses with large amounts of chirp or coding. The numbers of individual frequencies and their instantaneous bandwidths can be chosen from within this overall bandwidth. Digital waveform generation within the MAR allows it


to use adaptive waveform and frequency selection.

Array design Choice of elements and

spacing: The design of the array is a tradeoff between the EIRP required, the sidelobes and the scan volume required.

The scanning performance of the array is a function of the radiating element design and the element spacing. The elements need to be spaced such that when the beam is scanned to the maximum extent grating lobes are not generated. Grating lobes are a second representation of the main beam and occur if the element lattice is too large. If grating lobes occur the main beam gain is reduced and the radar may target Fig. 11 Adapted antenna pattern. Adaption with 12 subarrays and 5 jammers. positions in space. The appearance of grating lobes can be predicted. Fig. 9 is a contour plot of the radiation pattern of an antenna showing a grating lobe generated for a specific array lattice. In this illustration the beam is scanned far enough to cause the grating lobes to appear.

A second phenomenon, which needs to be assessed, is that of blind angles. Blind angles are a function of the array spacing, lattice geometry and the specific element design. At a blind angle the mutual coupling between elements results in the active reflection coefficient of the array approaching unity, the gain falling to zero with no radiation taking place. At a blind angle all transmitted power is reflected back into the active modules, possibly resulting in catastrophic failure of the system. The prediction of blind angles is not as straightforward as the prediction of grating lobes.

Besides avoiding grating lobes and blind angles the design must be such that sufficient EIRP is available at the required scan angles. The gain at a given scan angle is a function of the broadside aperture gain and the radiation pattern of the array element. This generally results in a loss of gain with scan angle that approximates to a COS^.^ or a cos2 function.

The broadside gain of the array is afunction of the array area and the amplitude taper applied in order to reduce the sidelobes. In traditional array designs the RF beamforming network applies the taper. In active arrays the transmit/receive active modules can be used as well, if required, to add a taper. The modules can be operated in class A on transmit and/or fitted with controllable

attenuators to apply the required receiver taper. Power and efficiency considerations, however, generally mean that the power stages operate in class C and no amplitude taper is applied on transmit. For large arrays with high numbers of elements the possibility exists to provide phase weighting to shape the beam.

As well as the design of the radiating aspects of the active antenna the other areas that require consideration are the receiverhransmit module and the mechanical and thermal design of the antenna.

TxRx module: This module contains the transmit power stage, low-noise receive amplifier and limiter, associated phase shifters, attenuators and circulators. Filtering must be provided to band-limit emissions and to provide protec- tion against out-of-band interference. Together with the microwave elements the module must also contain any control, communication and power conditioning electronics that are required. Generally modules are grouped into LRUs (Line Replaceable Units) containing a number of RF channels to optimise the use of silicon in the control electronics and the power conditioning components.

The modules must be housed, powered and cooled. The array structure carries out these functions. The cooling of the modules is particularly critical. In order to maintain the performance of the RF modules they must be held within a required temperature range. The design of the cooling system is seen as key to the performance of the array.

Subarrays: In order to carry out digital adaptive beamforming more than one receiver channel is required;


video

data ‘ IandQ

clutter v e I o c i ty

Fig. 12 AAPAR signal processing

in the limit, a receiver channel could be provided for each receiver module. Practical considerations, however, normally limit the number of receiver channels to the low tens. In order to do this the transmit/receive modules must be grouped into subarrays by the use of traditional RF beamforming techniques. The subarrays are generally randomised in shape, i.e. in height and width, but have a similar area. This improves the operation of the digital adaptive beamforming and breaks up any correlation effects.

Digital adaptive beamforming: A simplified layout for an adaptive beamformer is shown in Fig. 10. Each radiating element of the active array has its own low-noise amplifier (LNA) . Small groups (around 100) of co-located modules are combined in microwave networks to form subarrays (typically 8, 16 or 32). Each subarray is provided with a down-converter and a digitiser, which produces an accurate version of the amplitude and phase of the received signal.

The subarray channels can simply be summed to provide the normal ‘unadapted’ or quiescent antenna pattern, which would receive main beam target signals, with clutter and any noise jamming entering via sidelobes. In the adaptive beamformer each subarray receiver signal is adjusted in amplitude and phase (weighted) before summing to shape the radiation pattern.

The antenna pattern is modified so that nulls in the antenna sidelobe pattern are ‘driven’ into the direction of noise jammers. At the same time the main beam remains pointing at the target. Unlike some sidelobe canceller systems, the beamformer does not use any feedback and the signals appear at the output at the same time, as if the sum arrays where summed together, i.e. the nulls are formed at the same time as the main beam.

Several parameters can be adapted simultaneously and N-1 nulls can be driven into the array pattern, where N

is the number of adaptive channels. Fig. 11 shows an example of an antenna pattern adapted to null out five interference sources. Their position is marked as circles on the azimuth axis. In this case the adaption is taking place for both a sum and difference antenna pattern.

The beamformer can provide more than one output by processing the input signal in different ways. In addition to the standard sum output (used for surveillance and tracking), a monopulse and a sidelobe blanking beam can be provided. The monopulse output may be used to provide a two-dimensional measurement of the angle

offset of a target or own missile track from the boresight. This permits the absolute angular position to be output from the radar based on the known mechanical antenna position and the measured electrical boresight.

Signal generation The adaptation of the AAPAR is not limited to the

antenna beam patterns. The time management and waveforms of the radar must also be adapted to suit the radar’s various roles. This requires that the signal generator be capable of generating pulses of varying lengths, pulse repetition intervals (PRIs) , compression ratios and coding.

The signals to be transmitted for each radar task will, as in a conventional radar, generally use coded waveforms to achieve the desired performance. The use of frequency swept pulses for pulse compression purposes is well established and coding laws are available to meet various criteria. These criteria have to be changed adaptively as the role of the radar changes.

AAPARs, which derive low peak power, relatively high duty pulses from solid-state modules, use long pulses. This requires digital pulse compression and expansion techniques coupled with digital frequency synthesis under the control of the radar’s management system. Digital synthesis must be employed to achieve the very high stability needed to achieve the required clutter filtering and target Doppler filtering. The requirement to carry out target identification puts further demands on the stability and coherence of the signal source.

Signal processing Signal processing is a generic term used to describe the

filtering and extraction of data from radar signals. In common with the trends in conventional radars, MAR signal processing is increasingly carried out in software.


Fig. 13 Radar task scheduling process

The processes in an MAR are essentially the same as those in a conventional radar. The sequences in which these processes are performed are, however, much more complex because the radar performs multiple functions and can perform these in an almost random manner.

The signal processing function must be configured to accept signals in ‘batches’ that require specialised processing depending on the role or task that the signals represent. The radar management software has the task of controlling the processing to suit the current batch.

The processing carried out on each batch is familiar:

moving target filtering Doppler filtering integration background averaging plot extraction track extraction.

A typical arrangement of the processing is shown in Fig. 12, in which the process appears in the form of a time sequence. The individual subprocesses may in fact be performed by the same hardware components in non- real-time. This form of processing uses ‘conventional’ computing elements to carry the algorithmic transforms whilst storing intermediate data until required. The sequence in which the transforms are carried out may then be optimised to the workload of the signal processor.

Radar management In order to survey a given volume a number of beam

positions must be visited by the radar. The calculation of the number of beam positions is complex, being related to geometrical beam broadening effects*.

The degree to which the beams are required to be overlapped depends on the detection requirements. The number of transmit pulses required at each beam position is a function of the detection requirement and the required false alarm rate. These in turn are functions of the instrumented range, the size of the target to be detected and/or tracked, requirements for clutter filtering, etc.

The radar management system is designed to control and optimise the radar process to perform these tasks at the correct time (e.g. when a target is nearest to broadside). When peak loading causes short-term problems with radar resources, the manager is designed to act on task priorities, rescheduling tasks to maximise the value of the radar data to the defence system and making optimum use of the radar’s time x power product.

The radar management function has to co-ordinate the process of signal generation, beam pointing, dwell,

*These effects occur from two sources. When a beam is scanned off boresite, the effective aperture, when viewed from the beam direction, reduces, hence the beam broadens. Secondly, if the antenna is rotating then the higher the elevation of the target the wider the azimuth beamwidth; this is a result of the conical geometry of a rotating system.

ELECTRONICS & COMMUNICATION ENGINEERING JOURNAL, DECEMBER 2001 287

transmission, reception, signal processing and data extraction to ensure that the correct parameters are applied through each process to carry out the task demanded.

Typical processes used in task scheduling are depicted in Fig. 13.

Power and cooling Solid-state power modules require to be fed with a

source of prime electrical power capable of supplying the mean RF power requirements; peak power can in most cases be accommodated by local storage. If power is supplied at a voltage suitable for direct regulation down to the power stage then very-high-current power distribution networks are needed with the resulting problems of feed losses and voltage fluctuations. If high- voltage distribution is used then accessibility and safety become a problem. An adequate level of safety, however, can be achieved with a high-voltage distribution system. It is thus generally optimum to distribute power to the modules using a high-voltage distribution system and step-down at module or LRU level, with final regulation being provided as close as possible to the output devices.

To keep the array stable and to minimise the demands on any calibration system being used the active modules need to be cooled. The cooling needs to ensure that, under all environmental conditions, the active devices stay below a predetermined temperature and that the temperature gradients and fluctuations across the array face are not too large. The cooling problem gets much worse at higher frequencies due to the much higher waste power density generated at the array face, this being caused by the fall-off in RF efficiency coupled with the smaller lattice spacing. At L, S and C bands forced air-cooling is generally acceptable; for some C, X and higher bands liquid cooling may be the only option.

Integrated IFF IFF, which operates at the frequencies of 1030 and 1090

MHz can be integrated into the array by making use of the basic broadband nature of transmit/receive modules operating at L and S bands. At C and X bands other techniques must be adopted. This gives the possibility of selectively interrogating targets and performing immediate target confirmation interrogations. The use of relatively narrow elevation beams also gives rise to the possibility of using 3D monopulse IFF systems on mobile platforms.

Stealth, NCTI, LPI Waveforms have been developed that improve the

radar's ability to detect and track low-visibility (stealth) targets. Waveforms have also been developed that allow the radar to extract critical target features for use in a target classification process. This process is referred to as non-co-operative target identification ("I). It is also possible to adapt the waveforms to make the radar have a low probability of interception (LPI) by hostile ECM systems.

\ A- niques Group up until 1993. In 1993 he became a consultant to the ' Radar Systems Division, specialising in SSR system design and product saftey. During this period he was the System Design Authority for several SSR projects. Following the take-over of Siemens Plessey by BAe and the formation of BAE SYSTEMS he was appointed Head of Product Integrity and Safety for the Combat and Radar Systems (CaRS) business. In his present job, which covers all of the CaRS products, he still keeps an interest in radar system engineering issues. John is an IEE Fellow and a member of FEANI.

Address: BAE SYSTEMS (Combat and Radar Systems) Ltd., Newport Road, Cowes, Isle of Wight PO31 SPF, UK. E-mail: [email protected]

'

These techniques use wide bandwidths and extended target dwell times and are generally not suitable for use with the surveillance and tracking functions of the radar. The MAR, however, because of its infinite ability to adapt, can intersperse these more specialised modes of operation, depending on the radar's perceived role and the threats at any given time. Waveforms can be adapted on a pulse, burst, sector, scan or mission basis.

5 Summary

The MAR can provide many benefits in meeting the performance that will be required by tomorrow's radar systems. In some cases it will be the only possible solution. It provides the radar system designer with an almost infinite range of possibilities. This flexibility, however, needs to be treated with caution: the complexity of the system must not be allowed to grow such that it becomes uncontrolled and unstable. The MAR breaks down the conventional walls between the traditional system elements-antenna, transmitter, receiver etc.- such that the MAR design must be treated holistically. Strict requirements on the integrity of the system must be enforced. Rigorous techniques must be used to ensure that the overall flow down of requirements from top level is achieved and that testability of the requirements can be demonstrated under both quiescent and adaptive conditions.

Acknowledgment

This paper is based on the work carried out at BAE SYSTEMS CaRS (Combat and Radar Systems) over many years. The author wishes to thank the management of CaRS for permission to publish this paper.

OBAE SYSTEMS (Combat and Radar Systems) Ltd.: 2001 Received 11th September 2001


mailto:[email protected]

design considerations for adaptive active phased-array 'multifunction' radars

Documents