Radar Techniques Using Array Antennas is a thorough introduction to the possibilities of radar technology based on electronic steerable and active array antennas. Topics covered include array signal processing, array calibration, adaptive digital beamforming, adaptive monopulse, superresolution, pulse compression, sequential detection, target detection with long pulse series, space-time adaptive processing (STAP), moving target detection using synthetic aperture radar (SAR), target imaging, energy management and system parameter relations. The discussed methods are confirmed by simulation studies and experimental array systems developed by the authors team at FGAN, now Fraunhofer. This new edition has been fully updated and revised, and includes discussion of compressed sensing and its possible application to beam forming, some results for phase-only-nulling against jammers, descriptions of further algorithms for superresolution for location and separation of radar targets and the reconnaissance of other radiating sources, extension and explanation of the basic ideas for MIMO-radar, and a new chapter on radar operation by passive coherent location. Providing many valuable lessons for designers of future high standard multifunction radar systems for military and civil applications, this book will appeal to graduate level engineers, researchers, and managers in the field of radar, aviation and space technology.
Inspec keywords: pulse compression; radar antennas; radar detection; synthetic aperture radar; radar tracking; signal sampling; statistical analysis; target tracking; adaptive signal processing; MIMO radar; mathematical analysis; antenna arrays; array signal processing; signal representation; passive radar; signal classification
Other keywords: monopulse direction estimation; jammer suppression; polyphase code; active phased array; floodlight; inverse synthetic aperture radar; adaptive beamforming; mathematical tool; space-time adaptive processing; signal representation; statistical signal theory; array antenna; target detection; passive radar; array processing; MIMO radar; target classification; sequential detection; pulse compression; signal sampling; radar technique; signal digitisation
Subjects: Signal processing and detection; Radar equipment, systems and applications; Mathematical analysis; Antennas; Other topics in statistics; General electrical engineering topics
The most demanding requirements for radar systems have so far nearly always resulted from military objectives; civil applications then benefit from the achieved results. Developments in the available technology, particularly within the areas of highly integrated semiconductor technology for digital signal processors, miniaturised microwave integrated circuits (MMICs) and efficient microwave computeraided design methods, have enabled ever more demanding procedures and concepts to be achieved. The theoretical fundamentals and evaluation possibilities improved simultaneously, especially by the development of computers and programs. The use of powerful personal computers with program systems such as MATLAB or Mathematica is now very common.
In this book we will consider antenna arrays which consist of many individual antenna elements, and therefore a large number of signals, one for each element, have to be processed at the same time. The signals can be assigned to locations on the antenna, forming discrete sampling of the spatial wave field for transmitting or receiving. Signals transmitted or received with a radar system have also to be represented as a function of time. Signal samples are formed in the spatial and temporal dimension for digital signal processing with signal processors or computers and for recording for later analysis.
This chapter discusses the function of signal processing to detect the targets from the received signals as effectively as possible and in particular to extract afterwards further information for these targets, especially parameters such as amplitude, Doppler frequency shift, Doppler spectrum, polarisation and exact position. In some applications even a target image can be extracted.
In this chapter we will discuss the most important and well-known relations and parameters of an array antenna that are relevant for the conception of a radar system. Possible implementations for different applications will be summarised. We will start our discussion with a simple illustration, the basic principle of which is sketched in Figure 4.1. A set of antennas or an array of antenna elements is distributed on a metal ground plane, preferably on a regular grid. These antenna elements may be, for example, dipoles matched with the length of their arms to the operating wavelength l of the radar system. In the case of transmission each antenna element is the source of a spherical wavefront. As a first step we assume waves of equal phase from each antenna element. Comparatively, we observe circular water waves if a group of persons standing at a linear sea wall are throwing stones at the same instant of time into calm water. These waves superpose coherently according to the famous Huygens' Principle at each point in space. If all sources radiate in phase then at the boresight direction, orthogonal to the antenna plane, we have a linear or coherent summation of the field strength of all individual waves. That is if, at a certain distance from the antenna, the electrical field strength produced from one antenna element is E, we would have with N antenna elements the field strength NE resulting at the boresight direction, or a power density proportional to (NE)2=N2E2. Outside the boresight direction the condition of in-phase superposition is not fulfilled and there is approximately, as the spatial mean over all directions, only a superposition of the power; that is we have there a power density with an order of magnitude only proportional to NE2. More detailed considerations for the antenna pattern will follow. This results in a factor of N for the power density between the boresight direction and all other directions. In this example our main beam is formed in the boresight direction and in the other directions only the unwanted sidelobes are produced. We recognise that only for N approaching infinity can we expect the sidelobe to mainlobe power ratio to approach zero. That would require an antenna with an infinite antenna plane. The angular width of our main beam, the region with approximately coherent superposition of the partial waves from the antenna elements, also depends on N. The angular region for an in-phase superposition of all individual waves decreases for an increasing N, resulting in a narrower beam.
Beamforming with all antenna elements of an array for transmitting or receiving means reproducing a desired beam pattern as closely as possible. Generally, a narrow main beam with high gain at the centre in the desired direction combined with low sidelobes for all other directions is required. The fundamental procedure for beamforming is, weighting the individual signals of all antenna elements and then summing all these weighted signals.
At the output of beamforming and the receiver channel, with a band-pass filter at the intermediate frequency (IF), the received signals are available for further processing. They are at this stage continuous analogue signals. For digital processing we need sampled values of the orthogonal components I and Q as described in Chapter 2 with (2.1). In this chapter we will discuss the necessary sampling rate and methods to derive the orthogonal components I and Q.
This chapter discusses pulse compression with sufficient Doppler tolerance may be achieved using polyphase codes derived from linear and nonlinear frequency modulation.
Generally, a series of echo signals is the basis for target detection. An arbitrary length of this series is applicable by using the beam agility, which enables adaptation of the transmit energy to target strength and of the required Doppler resolution. The step-scan operation of a phased array provides an important difference when compared with mechanical scanning radars with respect to the detection procedure and clutter suppression. Clutter suppression may be divided between strong fixed clutter and varying Doppler-shifted clutter, the first of which may be suppressed by a simple recursive filter. Using post processing the latter can be arranged advantageously, in an adaptive manner, after the filter-bank. Sea clutter suppression can be improved by a special multiplex operation of the beam, resulting in elongated dwell times in each BP. For a longer series of pulses, a possible Doppler shift during the dwell time has to be considered. The usual filter bank shows a degraded detection performance because the target energy is distributed into several filters. An autocorrelation estimation test will give a more robust performance in this case. The detection performance is evaluated and compared on the basis of simulation studies.
Sequential detection should be applied for the search mode of phased-array radars. A range-dependent weighting has to be used to combine the test function from range cells for forming one test function for the respective beam position. Then the highest gain, expressed in SNR saving, is achieved at maximum range. The most effective detection performance is achieved with coherent signal processing.
Military radar applications must generally take into account hostile counter-measures in the form of irradiation of interference signals from jammers, which emit different waveforms occupying the frequency band of our own radar as an electronic countermeasure (ECM). Usually for the investigation of counter-measures against this threat the assumed interference waveform is noise. The signal-to-noise ratio for target echoes at the output of the radar receiver would be dramatically decreased by this interference and target detection would be impossible or a lot of false alarms generated. Therefore we have to develop adequate electronic counter countermeasures (ECCMs) to maintain the operation of our radar. One common technique is to spread our own frequency band by changing the operating frequency using a random pattern for the frequency selection. This technique is named frequency agility. We thus force the enemy also to spread their noise power over a wider frequency band and therefore the spectral power density of the jammer in our receiver is minimised.
After the detection of targets in the search mode there follows a target location with range and direction estimation which is as precise as possible. This procedure may in principle be performed with the same receiving data as already used for detection. Multifunction radar systems with phased arrays may apply an additional acquisition mode after target detection to confirm the first target detection and thereby cancel to a large extent false alarms produced by noise and interference. This radar task may be performed with increased transmit energy compared to that used for the search mode because there are only relatively fewer acquisition orders. Because of the higher signal-to-noise (SNR) for the acquisition mode the location accuracy will be improved. The waveform will be chosen for improved range resolution, as discussed in Chapter 7. After target acquisition the location function will also be applied to the target-tracking process.
Super-resolution is offered by an active receiving array and special signal-processing procedures as an additional and new capability. The necessary SNR will be available by approaching target formations. The procedure for super-resolution is derived from likelihood-estimation theory. The resulting parametric target model fitting algorithm, combined with stochastic approximation, is recommended. It is applicable to coherent target signals. One prominent application is resolving the multipath problem, especially for locating low-flying targets above the sea.
For a radar onboard flying platforms (aircraft, drone, satellite) the direction-dependant relative velocities to the ground scatterers cause corresponding Doppler frequency shifts of the clutter echoes. Against this broadened clutter spectrum one can apply no common filter, because one would suppress too broad a Doppler frequency range of possible targets and would limit target detection inadmissibly and unnecessarily. With an active receiving antenna array, however, a signal field with samples in the space and time dimension can be offered for optimal signal processing for the detection of moving targets, even if these targets are slowly moving. The antenna array elements may form a linear or a planar array. The antenna can be looking with its broadside orientation forward or sideways relative to the flight direction. For an onboard multifunction phased-array radar, the forward-looking case is particularly of interest.
Synthetic aperture radar (SAR) with coherent focusing has been proposed in 1953 by C. W. Sherwin. It allows imaging of the ground scene in all weather conditions with high resolution using a radar on board a flying platform or satellite. The first actual systems applied optical processing for image generation. Since then many developments have achieved finer resolution and faster processing of received SAR signals by digital processing. New and extended capabilities can be achieved by applying an active phasedarray radar for SAR operation, including array signal processing methods as discussed in the preceding chapters. Opportunities are particularly good for the task of detecting, locating and imaging moving objects. The suppression of jamming may also be achieved.
In Chapter 14, we have seen high-resolution images of the ground scene produced by a radar flying on an elevated platform and thereby forming a long synthetic aperture. This long aperture resulted in a very fine azimuth or cross-range resolution. For ground-based radar systems it is often also highly desirable to get images of detected targets, e.g. of the observed aircraft. This could be a valuable contribution to a classification of flying targets.
Classification and possibly identification of observed radar targets is a natural requirement for radar systems applied for air or maritime traffic control and defence. In all cases without target information, e.g. provided by the secondary surveillance radar or with identification friend or foe (IFF), the radar echo signal has to be evaluated. For air surveillance it is of great importance to distinguish as fast as possible between targets of interest, such as aeroplanes, and false targets like birds or flocks of bird. Due to this distinction the formation and processing of false tracks, time and power of computer and radar are saved.
Research work in the area of phased-array radar has been performed at the research institute FGAN-FFM (now Fraunhofer-FHR) since 1970. The phasedarray system ELRA (electronic steerable radar) served as an experimental basis. The first successful radar operation with the tracking of flying targets dates back to the year 1975. Since then many modifications and further developments corresponding to the advances in theoretical knowledge and in technology havebeen made in the areas of hardware and software. Most of the new concepts described in this book have already been implemented and tested in the ELRA system. The main effort has been concentrated in the areas of signal and data processing and system control.
In this chapter, first floodlight and CW radar with an OLPI waveform, which distributes its radiated energy in space and time as much as possible is disussed. In comparison to a conventional radar, the peak power density for a certain location is reduced by the product of the usual transmit gain and the inverse duty factor, i.e. together about 105-107. The protection would be especially effective, if several systems operate in the same area at the same time and acting as mutual decoys. And the basic ideas of a MIMO radar are finally discussed.
In this chapter we will discuss some aspects of the choice of some main systems' parameters and their relations for multifunction operation. We consider the observation of the air space to detect, locate and track flying targets, the classical task of a surveillance radar. The aim is to establish a track for each target as early as possible. We have discussed in several chapters the arbitrary movement of the agile beam, which is generally a pencil beam. There is the desire to use this freedom in the most intelligent and effective way, which means to achieve reliable tracking of incoming targets with adequate location accuracy. The tracks should be established at a maximum range with a minimum of mean power. Thereafter, the tracks should be maintained with an adequate tracking rate and power.
Radar operation is possible without any dedicated transmitter by using an illumination of opportunity for the observed area. Examples are analogue and digital television systems (DVB-T), digital audio broadcast (DAB) and various mobile communications base stations. For example, we use the widespread existence of the digital wireless communication system GSM (global system for mobile communication). The base stations are ubiquitous and represent illuminators that may be used silently without any permission. Air and ground surveillance of moving targets up to medium ranges may be achieved. The transmitted power is comparable to that of the omnidirectional low probability of intercept (OLPI) system described in Chapter 18. Therefore a similar detection range can be expected.