STAP techniques can provide simultaneous rejection of jamming and clutter in the airborne radar. Although, in the past, STAP has been considered mainly for SLAR applications, there has been growing interest in applying the technique to the non-side-looking radar geometries. In this chapter, we assess the benefits of using STAP for simultaneously suppressing the clutter and jamming in the forward-looking phased array geometries. We initially analyse the clutter suppression problem and show how significant the rejection of both mainlobe and sidelobe clutter is achieved using the appropriate STAP approaches. This chapter includes a brief consideration of the effects of variations in the platform orientation on clutter suppression and slow moving target detection. The effects of manoeuvre (e.g. platform yaw, pitch or roll) on clutter and jammer suppression are examined and the relative merits of various approaches for compensating for platform manoeuvre are also assessed.

A significant challenge for the effectiveness of STAP techniques against real data is presented by the operation against severe and non-homogeneous interference environments. However, it is impossible to use such a large number of degrees of freedom adaptively, since this would yield unacceptable adaptivity losses. Moreover, the real-time implementation requirements demand the use of filters with a low computational cost. In this chapter, we describe three possible solutions for robust and effective STAP of radar data such as (i) an adaptive two-dimensional FIR filter with small support, (ii) non-linear non-adaptive schemes, and (iii) non-linear combination of adaptive two-dimensional FIR filters.The performances are evaluated and compared by a theoretical analysis and by application to a set of recorded radar data.In summary, the non-linear adaptive detector promises a remarkable detection performance in a non-stationary clutter background containing interfering targets.

This chapter describes the application of STAP (space time adaptive processing) to the synthetic aperture radar (SAR) systems. SAR is a microwave sensor that allows a high resolution mapping of electromagnetic (EM) backscatter from an observed scene. A two-dimensional image is provided in the radar polar coordinates, i.e. slant range and azimuth. High resolution in the slant range is obtained by transmitting a coded waveform with a large value of the time-bandwidth product that coherently processes the echoes in a filter matched to the waveform. High resolution along the transversal direction is achieved by forming a synthetic aperture. This chapter describes in detail the multichannel SAR (MSAR).

This chapter focuses on the adaptive clutter suppression with sum (Σ) beam and difference (Δ) beams. It is assumed for this chapter that the spatial adaptive presuppression of jammers has been applied as a necessary prior to STAP. The most important feature of ΣΔ-STAP stems from the fact that antenna engineers have excelled in the design of low-sidelobe ΣΔ-beams, whether it is for a phased array or reflector-feed antenna.In fact, analogue beamforming techniques for ΣΔ-beams are so well developed that the more expensive digital beamforming method for ΣΔ-beams does not seem necessary. In other words, STAP with ΣΔ-beams may well be an affordable approach to high-performance airborne surveillance radars.

Airborne radar is an essential tool for military surveillance and reconnaissance. In this chapter, we discuss the possibilities of replacing the mechanically steered antenna by a phased array. The phased array antenna offers various advantages such as look agility through electronic beam steering, adaptive beamforming including jammer cancellation and superresolution, and slow target detection by STAP-based GMTI techniques. The chapter also compares several array architectures with 360° azimuthal coverage and STAP aptitude. Most of the presented concepts are based on circular dipoles.

RADARSAT-2 is a two-aperture SAR interferometer. When used for GMTI measurements, RADARSAT-2 uses beams in the 40° to 50° incidence angle range to maximise the radial velocity component of the vehicle motion. The airborne experimental SAR reported in this chapter is designed to replicate the RADARSAT-2 GMTI mode resolution and observation geometry and tests the data processing algorithms that can be migrated to the RADARSAT-2 GMTI processor. The greatest difference between the airborne and space-based SAR/GMTI capabilities arises from the relationship between the platform velocity and the along-track velocities of the moving targets. In the airborne case, the target speeds are a significant fraction of the radar speed and can reasonably estimate the azimuthal target speed.

Spaceborne radar (SBR) has been proposed for various military and civilian applications. To help mitigate the exorbitant cost of fielding and testing of the spaceborne radar prototypes, engineers have come to appreciate the key role that simulation technologies play in the development process. In this chapter, the author discusses the design of computer simulation tools suitable for modelling and evaluating the performance of spaceborne MTI radars that employ STAP techniques. This chapter firstly reviews the spaceborne MTI radar applications and radar design. This is then followed by a review of the STAP techniques typically considered for spaceborne MTI radar. With this background, the design of spaceborne radar (SBR) simulation tools is examined in detail.

Space-based radar (SBR) systems provide several important capabilities for the detection of moving targets that are not possible from an airborne platform. Although SBR systems have been studied for some time now, it is only recently with the rapid development of computational and antenna array technology that these systems have come closer to realisation. For this chapter, the authors focus on a notional SBR platform in the low-Earth orbit (LEO). First, notation and metrics are established for a moving target detection from an SBR with STAP followed by a discussion of the unique characteristics of ground clutter returns in SBR. In particular, the problem of range-ambiguous clutter is described. Next, the authors demonstrate the impact of range-ambiguous clutter on STAP performance and describe two means of overcoming the range-ambiguous clutter problem for pulse-Doppler waveforms, namely PRF diversity and an increase in elevation aperture. Lastly, the chapter considers the use of an alternative new waveform, namely a long single pulse waveform, that eliminates the ambiguities in both the range and Doppler of an SBR system.

This chapter describes the methodologies for online processing of the received radar data by a set of N antennas and M pulse repetition intervals (PRIs) for the calculation of space-time adaptive (STAP) filter output. The numerically robust and computationally efficient QR-decomposition (QRD) is used to derive the so-called MVDR (minimum variance distortionless response) and lattice algorithms. The novel inverse QRD (IQRD) is also applied to the MVDR problem. These algorithms are represented as systolic computational flow graphs. The MVDR is able to produce more than one adapted beams focused along different angular directions and Doppler frequencies in the radar surveillance volume. An analysis of the numerical robustness of the STAP computational schemes is presented when the CORDIC (coordinate rotation digital computer) algorithm is used to compute the QRD and the IQRD. Benchmarks on general purpose parallel computers and on a VLSI (very large scale integration) CORDIC board are also presented.

STAP-based radar systems must operate in heterogeneous clutter environments. In this chapter, the author has introduced the aspects of space-time clutter heterogeneity that describes the impact of clutter heterogeneity on STAP performance and briefly highlights some techniques to enhance the STAP implementation in heterogeneous clutter environments. The chapter introduces five basic classes of clutter heterogeneity and related effects: amplitude heterogeneity, spectral mismatch, CNR-induced spectral mismatch, moving (Doppler-shifted) discretes, and related effects such as mismatched angle-Doppler loci. Using a basic model in describing the space-time clutter characteristics, the chapter investigates the potential impact of each class of heterogeneity on detection performance. Using the measured airborne radar data, the chapter also corroborated a fundamental space-time clutter model and several heterogeneous clutter models.

A new method for the training of STAP adaptive weights has been introduced and demonstrated on the experimental data collected with an airborne sensor. The method uses a phase-based criterion, as well as power, to choose samples for STAP training and is considered as an improvement to power-selected training for post-Doppler STAP algorithms. The use of phase effectively removes the targets whose inclusion in the training results in target self-nulling and a degradation in clutter cancellation performance. The new technique was compared with the power-selected training on experimental data collected with an airborne radar using the PRI-staggered STAP algorithm. The performance improvement achieved over the power-selected training is significant as the target signals were properly excluded from the training set. In addition, the SINR loss estimated from the STAP applied to the experimental data was compared with a theoretically predicted SINR loss, showing good agreement over all Doppler frequencies.

In this chapter, the authors have presented four deterministic direct data domain least-squares techniques for nulling interferers and extracting the signal of interest. This approach generates the adaptive weights on a snapshot-by-snapshot basis that eliminates the need for auxiliary training data sets. These techniques are capable of handling both coherent and non-coherent interferers in a stationary or non-stationary environment. The four processors are the eigenvalue processor, and the three least-squares methods: forward, backward and forward-backward procedures which may be implemented in real time on a signal processing chip. After describing the radar environment used to demonstrate these techniques, the authors developed the one-dimensional (space domain) implementation for each of the methods. The chapter has shown the generation of system matrices using the spatial samples, solves for the adaptive weights, and estimates the amplitude of the desired signal. The performance of the forward, backward and forward-backward methods has been evaluated using the output SINR and adaptive weight pattern for each of the methods.

This chapter is a brief review of several traditional STAP schemes as well as an explanation of the complex detection environments in which many radar signal processors must operate. A review of the deterioration of the performance of algorithms, as well as the benefits and drawbacks of newly proposed adaptive processing approaches, has been provided. The chapter discusses that an engineering trade off exists between the design of the algorithm and its computational difficulties in real-world implementation. In the future, we look forward to the potential marriage of several of the presented signal processing techniques to provide overall system robustness and real-time target detection in real-world, non-ideal detection environments.

In this chapter the perspectives of STAP radar for application to ground target tracking have been pointed out. The role of the radial target velocity was put forward. If the radial velocity becomes zero, as in the case of tangential motion or a stop, the STAP processor cannot distinguish between the target and the ground clutter and, therefore, suppresses the target. Several degrading effects have been discussed, including the influence of a large bandwidth, jamming and radar ambiguities. A few preliminary thoughts on tracking convoys of vehicles have been added which might be useful to inspire further research in this field.

Ground surveillance aims at near real-time production of a dynamic ground picture. This task comprises track extraction and track maintenance of single ground moving vehicles and convoys, mobile weapon systems or military equipment, as well as low-flying targets such as helicopters. For long-range, wide area, all-weather and all-day ground surveillance operating at high data update rates, GMTI radar proves to be the sensor system of choice (GMTI: ground moving target indication). By using airborne sensor platforms in stand-off ground surveillance applications the effect of topographical screening is alleviated, thus extending the sensor's field of view. In this chapter, selected tracking aspects are discussed. The following topics are of particular interest: Doppler blindness, road map information, and sensor data fusion.

The chapter gives a fairly compact overview of some of the key ideas on broadband superresolution and interference suppression which are applicable for the spatial case of a multifunction radar. In particular for airborne radar, there is a great interest in providing a search and track radar with an additional target recognition feature by high range resolution or with an additional imaging mode based on SAR (synthetic aperture radar) processing. The typical feature of a multifunction radar is that it has a high gain antenna, which means an array antenna which consists of thousands of elements. The array elements are summed up in analogue technology into subarrays which are further processed digitally. The main interest of this chapter is to point out in which way can extend the usual narrowband processing with such a complicated array to broadband applications.

In this chapter, the use of spatio-temporal adaptive array processing in over-the-horizon radar application is considered in order to remove non-stationary multipath interference ('hot clutter'). Since the spatio-temporal properties of hot clutter cannot be assumed constant over the coherent processing interval, conventional adaptive techniques fail to provide effective hot clutter mitigation without simultaneously degrading the properties of the backscattered sea/terrain radar signals ('cold clutter'). The approach presented incorporates multiple stochastic constraints to achieve effective hot clutter suppression, while maintaining distortionless output cold clutter post-processing stationarity. The use of stochastically constrained spatial and spatio-temporal adaptive processing for hot clutter mitigation is discussed in scenarios that both do and do not allow access to a group of range cells that are free of cold clutter (supervised and unsupervised training, respectively). Theoretical and simulation results are complemented by surface-wave over-the-horizon data processing, collected during experimental trials in northern Australia. The final section discusses convergence rate issues for stochastically constrained adaptive algorithms based on loaded sample matrix inversion routines.

In this chapter, STAMP processing that combines STAP and MFP has been developed. Simulations show that STAMP coherently combines signal multipath spread in azimuth and Doppler and greatly enhances the target detection as well as providing target range and depth classification and localisation. In future studies, the robustness of STAMP against array shape error, frequency mismatch and environmental mismatch as well as how STAMP performs in other tactical scenarios will be addressed.

This chapter deals with reverberation suppression in active sonar systems. Reverberation is a phenomenon which is due to reflections of the transmitted acoustic signals by the ocean bottom, from the sea surface and from within the ocean volume. Reverberation limited environments like littoral waters are a severe problem for active sonar systems because the target echo level may not exceed the reverberation level. Active sonar systems are used in a sensitive marine environment. Designers and users of active sonar systems are aware of the potential threat of active sonar systems to the marine environment. This chapter presents sonar systems which follow rigorous mammal protection procedures in order to exclude any mammal injury due to active sonar operation.

This work deals with computationally efficient algorithms for estimating array signal parameters. In particular, the well known EM and SAGE algorithms and novel recursive versions have been studied intensively. Using data augmentation, the complicated multi-dimensional search involved in maximising likelihood functions has been simplified considerably. As SAGE has a more flexible augmentation scheme than EM, it often has a faster convergence speed. The fast version of EM and SAGE uses an adaptive procedure in the M-step to reduce parameter search spaces without affecting the estimation performance. The recursive EM and SAGE are stochastic approximation procedures with specialised gain matrices derived from the augmented information matrix. Because of the simple structure of the augmented data, the recursive EM and SAGE algorithms can easily be implemented in a STAP framework. Experimental results from sonar data have shown that the EM and SAGE algorithms provide similar estimates.

Space-time adaptive processing techniques play an important role in seismic reflection imaging. Their main applications are the suppression of coherent noise and random ambient noise in the recorded data to amplify primary reflection events, as well as obtaining kinematic information required for transforming the data into a depth image of subsurface structures. This chapter discusses a new data-driven processing technique, the common reflection surface task. The method makes use of data redundancy to obtain a simulated zero offset section with a significantly improved SNR while reducing the amount of data for further processing and imaging. In addition, kinematic information is extracted from data, which can be used in subsequent steps to construct a seismic velocity model, required for transforming the data into the depth domain.

In this chapter, linear JD techniques were presented as an application of STAP in the field of mobile radio communications. STAP was performed on the UL of a cellular SD/TD/CDMA system by adaptively equalising the multiuser system channel matrix, which contains the spatio-temporal channel impulse responses between all transmitting MS and the receiving BS antenna array and the code signatures of the MS. Perfect channel state information at the receiver was assumed, whereas in real-world systems blind/semi-blind/non-blind tracking of the spatio-temporal channel impulse responses has to be performed on a time slot basis. The same holds for the estimation of the spatial ICI covariance matrix. In particular, trade-off between deploying the available DoF in the space and code domain for exploitation of spatial and frequency diversity (to reduce fading) or for intra- and intercell interference cancellation (to create spatial transmission channels) was examined.

This chapter discusses space-time adaptive processing in the context of blind equalisation of underwater acoustic channels. To justify the great efforts in signal processing necessary to establish a reliable underwater communication link, special emphasis has been put on the characterisation of the transmission medium. The typical underwater acoustic channel has been found to be overspread due to rapid temporal and spectral variations, therefore calling for special signal processing techniques. Their development with time has been summarised in a brief history of underwater acoustic communication, highlighting only key contributions. Subsequently, a spatial-temporal receiver architecture has been introduced that efficiently allows for joint processing of signals received on many sensors. Based on this structure, a signal model suitable for the description of blind space-time adaptive equalisation algorithms has been developed. Then, the well known constant modulus algorithm for blind channel equalisation has been treated as an example for the class of stochastic gradient descent algorithms. The results obtained with measured shallow water communication data have demonstrated the general applicability of this algorithm but have also shown its slow convergence as a major drawback. As an alternative, an adaptive multichannel version of the Shalvi-Weinstein algorithm for blind equalisation has been derived. This algorithm is closely related to the constant modulus algorithm but shows a better convergence behaviour. This theoretically obtained result has been exemplarily verified by analysing the same data set as with the CMA. Although in this chapter we have presented two successful strategies for blind space-time adaptive equalisation with applications to underwater acoustic communication, blind equalisation is still an active area of research. In particular, for many proposed algorithms the bridge between theory and application is still lacking, constituting a rich set of research challenges for many years to come.

This chapter presents reduced-rank chip-level MMSE equalisers for the CDMA downlink with frequency-selective multipath based on the MSWF, for a known channel case and also for a training-based adaptation. The performance for the single base station case, and for the edge-of-cell scenario with soft hand-off are very satisfactory. The convergence rate for MSWF operating in a very low-rank subspace was significantly better than LMS, and somewhat better than RLS. The BER performance showed improvement over the full-rank methods for practical SNR range. This excellent performance is achieved at a computational complexity in between LMS and RLS due to the lattice-type structure which allows block-adaptive implementation through simple filtering operations. A space-time preprocessor algorithm was also presented based on the MSWF. A closed-form expression for the Wiener-Hopf weights based on the MSWF was derived which provides insight into the dependency of the Wiener-Hopf filter weights on the number of stages and blocking matrices at each stage. The preprocessor was shown to effectively null out both narrowband and wideband jammers while operating in a reduced-rank mode and minimising computational complexity.

This chapter concludes that smart antenna wireless communication systems provide significant gains in terms of spectral efficiency and link reliability. These benefits translate to wireless networks in the form of improved coverage and capacity. MIMO communication theory is an emerging area and full of challenging problems. Some promising research areas in the field of MIMO technology include channel estimation, new coding and modulation schemes, low complexity receivers, MIMO channel modelling and multiuser SDMA network design.