Unmanned aerial vehicle (UAV) has been widely used as an aerial base station (BS) to assist the ground communications for Internet of Things (IoT) due to its wide-area coverage, high mobility and line-of-sight (LoS) communication link. In this paper, a multi-UAV enabled IoT is considered, where the UAVs provide integrated sensing and communication (ISAC) services to the IoT nodes. The radar mutual information (MI) is introduced to measure the sensing performance of ISAC from the information theory perspective. To achieve fair communications, we seek to maximize the minimum communication rate per IoT node by jointly optimizing node scheduling, transmit power and 3D trajectory of the UAV under the constraint of radar MI for each IoT node. The formulated non-convex multi-variable optimization problem is divided into three subproblems, including UAV scheduling optimization, UAV transmit power optimization, and UAV 3D trajectory optimization. The near-optimal solutions of the original optimization problem can be achieved by proposing a three-layer iterative optimization algorithm to optimize the three subproblems iteratively. The simulation results demonstrate that the proposed optimization scheme can obviously improve communication rate under the constraint of radar MI as well as achieve fair communication for each node.

As an aerial service platform for Internet of Things (IoT), unmanned aerial vehicle (UAV) can provide integrated sensing, computing and communication (ISCAC) services for the IoT nodes. In this paper, a UAV-enabled ISCAC system is proposed for IoT to meet the evolving requirements of emerging services in 6G networks. This system has three functions: sensing user equipments (UEs) for acquiring radar sensing information, executing computing tasks, and offloading incomplete tasks to the access point (AP) for further processing. Through jointly optimizing UAV CPU frequency, UAV radar sensing power, transmit power of UEs, and UAV trajectory, the weighted total energy consumption of both the UAV and the UEs can be minimized. We present a three-layer iterative optimization algorithm to tackle the original non-convex optimization problem. Finally, the effectiveness of the algorithm and its superiority in energy consumption compared to other benchmark schemes are verified through simulation results.

Effective beam alignment is essential for vehicle-to-infrastructure (V2I) millimeter wave (mmWave) communication systems, particularly in high-mobility vehicle scenarios. This paper explores a three-dimensional (3D) vehicle environment and introduces a novel deep learning (DL)-based beam search method that incorporates an image-based coding (IBC) technique. The mmWave beam search is approached as an image processing problem based on situational awareness. We propose IBC to leverage the locations, sizes, and information of vehicles, and utilize convolutional neural network (CNN) to train the image dataset. Consequently, the optimal beam pair index(BPI)can be determined. Simulation results demonstrate that the proposed beam search method achieves satisfactory performance in terms of accuracy and robustness compared to conventional methods.

Recently, part information of pedestrian images has been demonstrated to be effective for person re-identification (ReID), but the part interaction is ignored when using Transformer to learn long-range dependencies. In this article, we propose a novel transformer network named Completed Part Transformer (CPT) for person ReID, where we design the part transformer layer to learn the completed part interaction. The part transformer layer includes the intra-part layer and the part-global layer, where they consider long-range dependencies from the aspects of the intra-part interaction and the part-global interaction, simultaneously. Furthermore, in order to overcome the limitation of fixed number of the patch tokens in the transformer layer, we propose the Adaptive Refined Tokens (ART) module to focus on learning the interaction between the informative patch tokens in the pedestrian image, which improves the discrimination of the pedestrian representation. Extensive experimental results on four person ReID datasets, i.e., MSMT17, Market1501, DukeMTMC-reID, and CUHK03, demonstrate that the proposed method achieves a new state-of-the-art performance, e.g., it achieves 68.0% mAP and 84.6% Rank-1 accuracy on MSMT17.

High-mobility unmanned aerial vehicles (UAVs) can serve as dual-function aerial service platforms for the Internet of Things (IoT), providing both sensing and communication services for IoT nodes without a base station (BS), particularly in emergency situations. In this paper, a UAV-assisted integrated sensing and communications (ISAC) system is proposed for IoT, which simultaneously senses the status information around the IoT and sends the sensing information to both the IoT nodes and a data collection center. In order to assess the sensing performance of ISAC, the radar estimation rate is introduced as a significant metric from the perspective of information theory. Considering the mutual interference between sensing and communications, the radar estimation rate is maximized through the coordinated optimization of UAV task scheduling, transmit power allocation, and 3D flight parameters under the constraint of communication rate. The formulated non-convex mixed-integer programming problem is divided into three subproblems, including UAV task scheduling optimization, UAV sensing and communication power optimization, and UAV 3D flight parameters optimization. The optimal solutions can be achieved by proposing a three-layer iterative optimization algorithm to optimize the three subproblems iteratively. The simulation results show that the radar estimation rate can well measure the sensing performance of the ISAC, which can be effectively improved by optimizing the 3D UAV flight parameters.

Abstract Cross‐modality person re‐identification is a challenging task due to the large visual appearance difference between RGB and infrared images. Existing studies mainly focus on learning local features and ignore the correlation between local features. In this paper, the Integration Graph Attention Network is proposed to learn the completed correlation between local features via the graph structure. To this end, the authors learn the coarse‐fine attention weights to aggregate the local features by considering local detail and global information. Furthermore, the Multi‐Centre Constrained Loss is proposed to optimise the feature similarity by constraining the centres of modality and identity. It simultaneously utilises three kinds of centre constraints, that is intra‐identity centre constraint, modality centre constraint, and inter‐identity centre constraint, in order to reduce the influence of modality information explicitly. The proposed method is evaluated on two standard benchmark datasets, that is SYSU‐MM01 and RegDB, and the results demonstrate that the authors’ method achieves better performance than the state‐of‐the‐art methods, for example, surpassing NFS by 4.8% and 6.0% mAP on the single‐shot setting in All‐search and Indoor‐search modes, respectively.

Abstract Cross‐modality person re‐identification (Re‐ID) aims to retrieve a query identity from red, green, blue (RGB) images or infrared (IR) images. Many approaches have been proposed to reduce the distribution gap between RGB modality and IR modality. However, they ignore the valuable collaborative relationship between RGB modality and IR modality. Hybrid Mutual Learning (HML) for cross‐modality person Re‐ID is proposed, which builds the collaborative relationship by using mutual learning from the aspects of local features and triplet relation. Specifically, HML contains local‐mean mutual learning and triplet mutual learning where they focus on transferring local representational knowledge and structural geometry knowledge so as to reduce the gap between RGB modality and IR modality. Furthermore, Hierarchical Attention Aggregation is proposed to fuse local feature maps and local feature vectors to enrich the information of the classifier input. Extensive experiments on two commonly used data sets, that is, SYSU‐MM01 and RegDB verify the effectiveness of the proposed method.

Industrial Internet of Things (IIoT) usually utilizes 2.4-GHz unlicensed frequency band, which is also heavily used by many other communication systems, such as ZigBee, WiFi, Bluetooth, etc. Therefore, the lack of spectrum resources has become a key technical bottleneck to restrict the development of IIoT. Integrating cognitive radio (CR) into IIoT, Cognitive IIoT (CIIoT) can cope with the spectrum resource shortage by accessing the frequency bands licensed to primary user (PU). However, spectrum sensing and access control must be performed to avoid bringing severe interference to the PU. In this article, an integrated cooperative spectrum sensing (CSS) and access control model is proposed to improve the transmission performance of the CIIoT while guaranteeing the CSS’s detection probability and controlling the interference to the PU. This model is optimized to maximize the total throughput of IIoT in each frame by jointly optimizing sensing time, the number of sensing nodes and the transmit power for each node under the constraints of the minimum detection probability, the total power control, the interference control, and the minimum rate for each node. The optimization problem is solved by the joint optimization of spectrum sensing and access control. A simultaneous CSS and access control model is also proposed to increase the communication time by using one time slot to perform CSS and access control simultaneously. The simulation results show that there exist optimal sensing and control parameters to maximize the total throughput of CIIoT.

Recently, convolutional neural network (CNN) dominates the ground-based cloud image segmentation task, but disregards the learning of long-range dependencies due to the limited size of filters. Although Transformer-based methods could overcome this limitation, they only learn long-range dependencies at a single scale, hence failing to capture multi-scale information of cloud image. The multi-scale information is beneficial to ground-based cloud image segmentation, because the features from small scales tend to extract detailed information while features from large scales have the ability to learn global information. In this paper, we propose a novel deep network named Integration Transformer (InTransformer), which builds long-range dependencies from different scales. To this end, we propose the Hybrid Multi-head Transformer Block (HMTB) to learn multi-scale long-range dependencies, and hybridize CNN and HMTB as the encoder at different scales. The proposed InTransformer hybridizes CNN and Transformer as the encoder to extract multi-scale representations, which learns both local information and long-range dependencies with different scales. Meanwhile, in order to fuse the patch tokens with different scales, we propose Mutual Cross-Attention Module (MCAM) for the decoder of InTransformer which could adequately interact multi-scale patch tokens in a bidirectional way. We have conducted a series of experiments on large ground-based cloud detection database TLCDD and SWIMSEG. The experimental results show that the performance of our method outperforms other methods, proving the effectiveness of the proposed InTransformer.

Person re-identification (Re-ID) is a challenging task due to variations in pedestrian images, especially in cross-domain scenarios. The existing cross-domain person Re-ID approaches extract the feature from single pedestrian image, but they ignore the correlations among pedestrian images. In this paper, we propose Heterogeneous Convolutional Network (HCN) for cross-domain person Re-ID, which learns the appearance information of pedestrian images and the correlations among pedestrian images simultaneously. To this end, we first utilize Convolutional Neural Network (CNN) to extract the appearance features for pedestrian images. Then we construct a graph in the target dataset where the appearance features are treated as the nodes and the similarity represents the linkage between the nodes. Afterwards, we propose Dual Graph Convolution (DGConv) to explicitly learn the correlation information from the similar and dissimilar samples, which could avoid the over-smoothing caused by the fully connected graph. Furthermore, we design HCN as a multi-branch structure to mine the structural information of pedestrians. We conduct extensive evaluations for HCN on three datasets, i.e. Market-1501, DukeMTMC-reID and MSMT17, and the results demonstrate that HCN is superior to the state-of-the-art methods.

Most ground-based remote sensing cloud classification methods focus on learning representation features for cloud images while ignoring the correlations among cloud images. Recently, graph convolutional network (GCN) is applied to provide the correlations for ground-based remote sensing cloud classification, in which the graph convolutional layer aggregates information from the connected nodes of graph in a weighted way. However, the weights assigned by GCN cannot reflect the importance of connected nodes precisely, which declines the discrimination of the aggregated features (AFs). To overcome the limitation, in this article, we propose the context graph attention network (CGAT) for ground-based remote sensing cloud classification. Specifically, the context graph attention layer (CGA layer) of CGAT is proposed to learn the context attention coefficients (CACs) and obtain the AFs of nodes based on the CACs. We compute the CACs not only considering the two connected nodes but also their neighborhood nodes in order to stabilize the aggregation process. In addition, we propose to utilize two different transformation matrices to transform the node and its connected nodes into new feature spaces, which could enhance the discrimination of AFs. We concatenate the AFs with the deep features (DFs) as final representations for cloud classification. Since existing ground-based cloud data sets (GCDs) have limited cloud images, we release a new data set named GCD that is the largest one for ground-based cloud classification. We conduct a series of experiments on GCD, and the experimental results verify the effectiveness of CGAT.

Many methods for ground-based remote sensing cloud detection learn representation features using the encoder–decoder structure. However, they only consider the information from single scale, which leads to incomplete feature extraction. In this article, we propose a novel deep network named dual pyramid network (DPNet) for ground-based remote sensing cloud detection, which possesses an encoder–decoder structure with dual pyramid pooling module (DPPM). Specifically, we process the feature maps of different scales in the encoder through dual pyramid pooling. Then, we fuse the outputs of the dual pyramid pooling in the same pyramid level using the attention fusion. Furthermore, we propose the encoder–decoder constraint (EDC) to relieve information loss in the process of encoding and decoding. It constrains the values and the gradients of probability maps from the encoder and the decoder to be consistent. Since the number of cloud images in the publicly available databases for ground-based remote sensing cloud detection is limited, we release the TJNU Large-scale Cloud Detection Database (TLCDD) that is the largest database in this field. We conduct a series of experiments on TLCDD, and the experimental results verify the effectiveness of the proposed method.

In this paper, we consider a reconfigurable intelligent surface (RIS) assisted unmanned aerial vehicle (UAV) relaying communication system, where the RIS is mounted on the UAV and can move at a high speed. Compared with the conventional static RIS, better performance and more flexibility can be achieved with the assistance of the mobile UAV. We maximize the average downlink throughput by jointly optimizing the UAV trajectory, RIS passive beamforming and source power allocation for each time slot. The formulated non-convex optimization problem is decomposed into three subproblems: passive beamforming optimization, trajectory optimization and power allocation optimization. An alternating iterative optimization algorithm of the three subproblems is proposed to achieve the suboptimal solutions. The numerical results indicate that the RIS-UAV relaying communication system with trajectory optimization can get higher throughput.

Sixth-generation (6G) wireless communications and networks will continue to move to higher frequency and wider bandwidth, with much higher data rate and spectral efficiency. Given the heterogeneity and densification of Internet of Things (IoT), 6G wireless network may need to be extended to modern random access (RA) for IoT applications, which can be achieved via smart protocol design and advanced signal processing and communications technologies. The modern RA techniques, such as massive multiple-input–multiple-output (MIMO), OFDMA, nonorthogonal multiple access (NOMA), sparse signal processing, or new orthogonal design techniques provide good candidacy for 6G-enabled IoT. In 6G, grant-free transmission should be designed for distributed IoT applications. IoT applications are often involved in self-organizing decision making. 6G-enabled IoT will take advantage of the recent development of artificial intelligence (AI) techniques. It will generate new knowledge and understanding and accelerate discovery and innovation in IoT. Each of these efforts is designed to amplify the intrinsically multidisciplinary nature of the emerging field of IoT. The 6G-enabled IoT will establish theoretical, technical, and ethical frameworks that will be applied to tackle many challenges in IoT, advancing technology for humanity.

Unsupervised person reidentification has received more attention due to its wide real-world applications. In this paper, we propose a novel method named fuzzy multilayer clustering (FMC) for unsupervised person reidentification. The proposed FMC learns a new feature space using a multilayer perceptron for clustering in order to overcome the influence of complex pedestrian images. Meanwhile, the proposed FMC generates fuzzy labels for unlabeled pedestrian images, which simultaneously considers the membership degree and the similarity between the sample and each cluster. We further propose the fuzzy label regularization (FLR) to train the convolutional neural network (CNN) using pedestrian images with fuzzy labels in a supervised manner. The proposed FLR could regularize the CNN training process and reduce the risk of overfitting. The effectiveness of our method is validated on three large-scale person reidentification databases, i.e., Market-1501, DukeMTMC-reID, and CUHK03.

Recently, multimodal cloud samples are utilized to learn completed feature representations for cloud classification. However, the existing methods neglect the related information from other multimodal cloud samples in the learning process, which leads to inadequate learning. In this article, we propose a novel deep model to learn heterogeneous deep features (HDFs) for multimodal ground-based remote sensing cloud classification. Specifically, we first design the convolutional neural network (CNN) extractor to combine the visual information and the multimodal information (MI) to obtain the CNN-based features of multimodal cloud samples. Afterward, we treat the CNN-based features of multimodal cloud samples as the nodes of graph, and utilize the similarity between nodes as the adjacency matrix. We feed the graph and the adjacency matrix into the graph convolutional network (GCN) extractor to obtain the GCN-based features that could capture correlations among multimodal cloud samples using graph convolutional layers. After obtaining CNN-based features and GCN-based features, we concatenate the two kinds of heterogeneous features to represent the multimodal cloud samples. As a result, the concatenated feature contains the visual information, the MI and the related information among multimodal cloud samples. We conduct a series of experiments on the multimodal ground-based cloud database (MGCD), and the experimental results verify that the proposed HDF outperforms state-of-the-art methods.

The Internet of Things (IoT) provides enormous connections of devices and sensors with different applications. It is an enabling technology for smart city, intelligent transportation systems, environmental monitoring, security surveillance, smart homes, satellite and space information network, ocean monitoring, and unmanned border awareness systems, just to name a few. IoT as a high-density network will take the burden of massive data generated by different kinds of terminals and sensors. Dramatic growth in IoT has created a shortage in the available radio spectrum. Wireless communications services in IoT such as cellular phones, tablets, and wireless Internet access have to compete with existing users in radar, government and military communications, environmental monitoring, and other IoT applications. The strategy to increase the efficiency of spectrum sharing among the enormous users in IoT. Besides, the IoT applications demand more and better functionality and performance from new electronic devices; these demands translate into greater energy consumption demands. The gap between energy storage and demand continues to grow and the battery technologies for energy storage are not expected to increase tremendously in the coming years. Furthermore, reducing signal transmission power can lessen interference among devices in IoT. Energy-efficient protocols and network architectures will further reduce the number of transmissions and extend the battery life of IoT devices (IoTDs). Therefore, it is essential to pursue fundamental research on new components, techniques, and architectures to achieve energy-efficient sensing, communications, and networking in a shared spectrum environment for IoT.

Internet of Things is coming to critical missions such as battlefield, border patrol, search and rescue, critical structure monitoring and surveillance, etc. Internet of Things (IoT) in Critical Missions or Internet of Mission-Critical Things (IoMCT) is propelled by the convergence of sensing, communication, computing, and control. To support IoMCT, the mission-critical networks will need to be flexible and interactive, and still work despite limited bandwidth, intermittent connectivity and with a large number of devices on the network. The focus of IoMCT is to improve surveillance utilizing a network, not fusion of disparate sensor products. Such adaptation, management, and re-organization of information sources, devices, and networks must be accomplished almost entirely autonomously, in order to avoid imposing additional burdens on the humans, and without much reliance on support and maintenance services.

The prompt delivery of biomedical images for emergency diagnosis purpose is an important issue in health care organizations. This paper is aimed a developing class of algorithms to view and manipulate medical images on mobile devices and mainly PDA handhelds. We illustrate our method on human brain scans to view: 2-D single medical imaging scans, multi frames/slices medical imaging scans, internal 3-D anatomical details of a simulated straight line-cut, and the reconstruction of the original scanned object e.g. the original head image.

This paper is concerned with a comparative assessment of policies and strategies currently being employed by a number of nations to promote science, technology and innovation as a means for wealth creation and prosperity. The work includes a study of policies implemented by the governments of the UK, Canada, USA, Germany and China, and an evaluation based on performance indicators developed by the EU commission's innovation scoreboard.

A new type of divergence measure for the registration of medical images is introduced that exploits the properties of the modified Bessel functions of the second kind. The properties of the proposed divergence coefficient are analysed and compared with those of the classic measures, including Kullback-Leibler, Renyi, and Iinfinity, divergences. To ensure its effectiveness and widespread applicability to any arbitrary set of data types, the performance of the new measure is analysed for Gaussian, exponential, and other advanced probability density functions. The results verify its robustness. Finally, the new divergence measure is used in the registration of CT to MR medical images to validate the improvement in registration accuracy.

This paper proposes the use of a polynomial interpolator structure (based on Horner's scheme) which is efficiently realizable in hardware, for high-quality geometric transformation of two- and three-dimensional images. Polynomial-based interpolators such as cubic B-splines and optimal interpolators of shortest support are shown to be exactly implementable in the Horner structure framework. This structure suggests a hardware/software partition which can lead to efficient implementations for multidimensional interpolation.

In this paper a novel application of the simultaneous perturbation stochastic approximation algorithm (SPSA) to the noisy non-stationary blind source separation problem is presented and described. The proposed approach demonstrates the algorithm with a second order cost function suitable for applications to non-stationary data. Some extensions to the algorithm that are currently being investigated are also described in the paper, and the algorithm performance is demonstrated via simulation.

We present an algorithm for performing attributed graph matching. This algorithm is derived from a generalized framework for describing functionally expanded interpolators which is based on the theory of reproducing kernel Hilbert spaces (RKHS). The algorithm incorporates a general approach to a wide class of graph matching problems based on attributed graphs, allowing the structure of the graphs to be based on multiple sets of attributes. No assumption is made about the adjacency structure of the graphs to be matched.

Blind synchronization of minimum-shift-keying signals in the context of time-selective fading communication channels is considered. Strictly feedforward algorithms based on second- and fourth-order cyclic statistics are proposed for frequency and timing offset estimation. The new estimators are shown to outperform existing methods. Extensions of the algorithm to GMSK are also studied. Finally, fourth-order cyclic statistics are shown to provide an accurate estimate of the Doppler spread.

This paper presents a new technique which exploits constrained optimization methods to derive optimal two dimensional filters in the cumulant domain for processing signals in non-Gaussian noise, or signals with corrupting interferences which have non-symmetrical probability density functions. The approach proposed for enhancing signals in such noise is important, as increasingly practical engineering application areas are identifying occasions where the perceived wisdom of modelling signals in additive Gaussian noise simply does not hold. Since the bispectrum of non-Gaussian noise and interference is not zero, it corrupts the bispectrum of the signal. Thus filters that suppress the bispectral component of the noise and enhance the signal bispectrum, are required. The two dimensional filters proposed have the property of concentrating the filter energy into a hexagonal region in the bispectral domain. This leads to an impulse response for these filters which represents a new form of two dimensional discrete prolate spheroidal sequence. The sensitivity of cumulant determination to non-Gaussian noise has been noted in the area of array processing. However this paper presents one of the first attempts to remove non-Gaussian noise by cumulant filtering.

Deterministic multiscale defined representational forms have found a significant role in the theory and application of signal processing. The most widely important form for signal and system modelling is likely to be multiscale defined wavelets. Another class of multiscale representation which has attracted consistent interest is the group of signal models defined in terms of iterated function systems (IFS). This paper is concerned with widening the IFS application to include system modelling, particularly of neural network-like structures. We introduce an interpolating IFS model as a form of self-organising map with global fractal constraints. Symbolic addressing is employed to discretize the attractor into pseudo-network nodes. We present in detail an online gradient based algorithm for training. This particular model is intended for efficient pattern recognition in complex environments, for example, with multifractal sources such as those seen in network traffic and general turbulence.

MHEG represents a new multimedia and hypermedia standard proposed by ISO/IEC. This paper presents a new software authoring environment based around MHEG-5 that offers users a vehicle for creating multimedia applications that can interact with external programs which involve intense computational tasks. MEMIS provides a linkage between a multimedia front-end and externally available computational processes. The paper provides a background to the development of the environment, by identifying the facilities offered by MHEG, discusses the efficacy offered by MHEG vs JAVA for multimedia development; and then covers the development process and includes specific exemplars of the environment for managing multimedia applications which include (a) real-time signal processing embedded within a LabVIEW kernel, (b) automatic teller machines, (c) set top box, and (d) kiosks for Internet commerce that utilises MATLAB type calls. The results include system level architecture for multimedia implementation, and the timing requirements for such applications.

A new approach for the decision feedback equalizer (DFE) based on the functional-link neural network is described. The structure is applied to the problem of adaptive equalization in the presence of intersymbol interference (ISI), additive white Gaussian noise, and co-channel interference (CCI). It is shown through simulation results for a severe amplitude distorted co-channel system that the decision feedback functional-link equalizer (DFFLE) provides significantly superior bit-error rate (BER) performance characteristics compared to the conventional DFE, the linear transversal equalizer (LTE), the nonlinear radial basis function (RBF) neural-network-based structures and the feed-forward functional-link equalizer (FFLE)-based structures. The DFFLE is also shown to have a significantly simpler computational requirement relative to the RBF and the FFLE.

There has been much interest recently in alternative methods of providing spread spectrum communications through the use of novel modulating carriers such as nonlinear chaotic signals, and localized wavelet and multirate filters. This paper presents research extending recent studies undertaken by the authors into the provision of CDMA spread spectrum through the use of iterated function system (IFS) or alternatively, multiscale difference equation (MSDE) defined pseudo-noise (PN) functions. While such equations are already important in the construction of wavelets and other models for signal coding and representation, here we are concerned with their ability to construct pseudo-random functions for use as broadband CDMA modulation signatures. Concentrating on fractal interpolation functions (FIF), we show how their construction can be written in terms of a Markov shift system and use this to qualify a proposed whitening procedure, and apply them as CDMA signatures. We subsequently use these functions in the construction of PN scaling functions and orthogonal wavelets.

An adaptive nonlinear filter based on a second order Volterra series and on an IIR filter structure is presented. This filter is able to model higher than second order nonlinearities for systems where the nonlinearities are harmonically related. This solution represents an alternative to using higher than second order Volterra filters. We present a full derivation of this gradient search based adaptive nonlinear filter and also highlight the various assumptions and simplifications which require to be made in order to produce a practical algorithm. A comparison is made in terms of the performance and computational complexity between an adaptive second order IIR Volterra filter and an adaptive second and third order Volterra filters.

In this letter, an adaptive recursive nonlinear filter based on the Volterra series and an infinite impulse response (IIR) structure is considered. For certain types of nonlinear systems where high-order nonlinearities are recursively generated, we show that the adaptive recursive second-order polynomial filter has improved performance over the well-known (nonrecursive) adaptive second-order Volterra filter and a third-order Volterra filter. This filter represents an alternative to using a traditional Volterra filter whose order has been increased to match that of the system being modeled.

Eigendecomposition based techniques such as MUSIC and its variants constitute effective methods for determining the direction of arrival (DOA) estimates of narrowband sources. A new strategy which extends the MUSIC algorithm to higher order statistics (HOS) is proposed for estimation of the DOA. Also, we present a new method for the estimation of the number of multiple narrowband incoherent and coherent non-Gaussian source signals arriving on the array which we consider as a significant contribution. The performance of the technique is compared with other suggested HOS-based methods.

Modern manufacturing systems call for full-rate automated inspection of produced samples. A vision-based intelligent perception system (IPS) is presented making automated inspection of chicken meat feasible. The IPS analyzes RGB images framing the chickens after the slaughtering and plucking process and detects defects such as burns, hematomas and blisters, along with other relevant features. The vision module, which is the core of the system, operates by first extracting the chicken body from the background, then it segments the body into its anatomic subparts. Defective areas are identified by means of morphological reconstruction. Finally, defects are classified by comparing their features against the defect description contained in a reference database.

This paper presents a background to the teaching of Signal Processing in Scottish universities, identifies commonality of undergraduate teaching programmes and contents, provides information on the key teaching syllabuses and on the relevant advanced final year options, as well as on postgraduate instructional courses. It also comments on current trends and future directions, and ends by posing a number of questions for educators.>

This paper brings together two strands of current interest in signal processing. While second order techniques and minimum variance criteria are well understood, there is a growing requirement to study the performance criteria that involve higher order statistics in order to evaluate deviation from Gaussianity, linearity and stationarity of observed data. There is a complimentary requirement for the generation of random test sequences which have prescribed (or minimal) higher order statistics, to facilitate the analysis of systems in order to determine linearity/non-linearity, time invariance vs time varying parameters. This paper proposes a new method for minimising the third order cumulant spread of random sequences with symmetric pdf, and provides a closed form solution for the weightings required to achieve this. Numerous computed results are included to verify performance.>

This paper presents a study of a new class of adaptive channel equaliser, known as the functional-link (FL) equaliser which utilises functional expansion of the equaliser's input data. By carefully selecting the appropriate functions of the input, significant performance improvement can be obtained. These studies will lead to significant reduction in the exponentially increased expansion of the polynomial-perceptron equaliser. Numerical simulation results are presented to highlight the better bit error rate performance of the FL based equaliser compared to other non-linear, neural network based equalisers.>

Novel wavelet transform based schemes for coding still images and image sequences are presented. The still image codec uses a new efficient adaptive bit-plane run-length coding of the wavelet transform coefficients of images. The main merit of this coding scheme is its simplicity requiring no training or storage of codebooks and it outperforms the JPEG at low bit-rate. For image sequence coding, a very low bit-rate sub-band motion estimation/compensation video codec designed for colour video conferencing applications is presented. A full system with buffer feedback control is designed. Simulation results of transmitting colour image sequences at 9.6 to 19.2 kbit/s are given.>

The design for a DSP (digital signal processing) subsystem for a health monitoring system for gas turbine engines is described. Knowledge-based techniques are emerging as useful tools for health and condition monitoring of high value engineering systems. A demonstrator system that uses these techniques for monitoring the health and performance of a marine gas turbine engine is described, with particular emphasis on the DSP subsystem which interfaces directly with the raw sensor data coming from the monitored system. The DSP subsystem is a coupled system using both numerical and symbolic methods for signal interpretation; i.e., the DSP subsystem provides a description of the sensors signals in meaningful symbolic terms that reflect the state of the monitored system. The signal abstractions are put in a form suitable for symbolic processing by the knowledge-based diagnostic subsystem of the monitoring system to determine the health of the monitored system.>

The stability of the noncanonical least mean square (NCLMS) algorithm is investigated. The NCLMS effectively uses a different step size for each tap coefficient position during adaptation. The classical LMS step size bound cannot be directly applied to the NCLMS. The weight error vector is modeled as a first-order difference equation and a stability bound for the NCLMS is derived. Simulation results are presented to back up the analysis.>

In seismic data interpretation, horizon picking is important for structural analysis, feature recognition, and site appraisal. However, horizon picking is still commonly done by hand, a process which is error prone and time consuming. Attempts to automate horizon picking are hindered by the absence of a clear, robust, and universal picking algorithm. A new method which combines a traditional approach to horizon picking with a new technique using a trained artificial neural network is presented. It is shown that this method makes better use of the general properties of horizons, is more robust than conventional pattern recognition techniques, and facilitates a solution to the problem of tracking through conventionally difficult regions containing faulting and other geophysical anomalies, where horizons are discontinuous.>

The use of iterated function system (IFS) and recurrent iterated function system (RIFS) representations of Markov and hidden Markov processes and dynamical systems as time series models is investigated. Originally devised for use in deterministic fractal image compression schemes, they are introduced as a time-varying, generally nonlinear stochastic data models which entail a new set of inverse problems. It is shown how IFSs, in their most general formulation, offer a framework which, in addition to describing a number of existing Markov models, suggests a range of diverse new ones. Emphasis is placed on a novel model whose RIFS version can be seen as a generalization of hidden Markov models having a greater predictive ability. Parameterization is discussed from both the synthesis and analytical viewpoints. An algorithm for the parameter estimation of the RIFS model is proposed. Comments are made on the relation between RIFS and deterministic chaotic nonlinear systems.>

The authors present analysis and simulations of an LMS (least mean square) based adaptive filtering algorithm called the NCLMS (non-canonical LMS). Rather than using the standard FIR (finite impulse response) filter as for the LMS algorithm, a modified structure called the NCFIR (non-canonical FIR) is used. The NCFIR allows a faster VLSI implementation than the conventional FIR. A comparison of the performances of the NCLMS and conventional LMS algorithm is presented for an inverse system modeling application. Simulation results are given which show a reduced EMSE (excess mean square error) level and an improved performance in an impulsive noise environment for the NCLMS over the LMS algorithm.>

A computationally efficient algorithm for direction finding is described. The sample covariance matrix is averaged into d column vectors, where d is the known number of sources. These column vectors are used as a basis for the signal subspace. They can be used directly for the two subspace estimates in the ESPRIT algorithm or they can be used to construct a projection matrix for use with the MUSIC algorithm. For low signal-noise-ratio conditions, a subtraction process is used to remove the diagonal noise terms from the covariance matrix and as a result a shorter search vector is used in the MUSIC algorithm.>

A new reconstruction technique is described to obtain the interior temperature field of composites from the reflected surface transient time-varying temperature response, or reflection thermogram. It is shown that the temperature field at an imaging time (t=0) represents the cross-section of the composite, which may be used to determine the presence, or otherwise, of any manufacturing flaws in the material. The generation of simulated reflection thermograms (forward problem), using a finite difference implicit method, is presented first. This method is then verified experimentally by comparing simulations with recorded data. The reconstruction technique (inverse problem) is then developed using theoretical principles derived for seismic migration algorithms. It is shown that the reflection thermogram migration algorithm involves forward extrapolation of the thermogram followed by an imaging step.>

The maximum likelihood estimator is the optimal estimator of the direction of sources, but it requires the minimization of a complex, multimodal, multidimensional cost function. A neural optimization procedure is presented that does not require an initial estimate of the direction of the sources and offers the potential of real-time solutions to the direction of arrival problem by utilizing the fast relaxation properties of the Hopfield network. A modification based on an iterated descent procedure is introduced into the Hopfield model dynamic equation to increase the probability of convergence to the global optimum. The algorithms are implemented on an array of closely coupled transputers that perform the random asynchronous neural updates in parallel. The mapping is achieved using a technique called chaotic relaxation. Simulation results are presented to characterize the performance of the neural approach in terms of the variance of the estimates of source directions and the time required for the computation of the estimates.>

A parallel implementation of the two-dimensional, adaptive blocksize discrete cosine transform (DCT) is based on the principle of the quadtree structure. An efficient coding technique for the quadtree used is described. A network of Inmos T800 transputers is chosen as the target system. Two types of parallel network topologies are analyzed, and the results for execution time are presented. The results show that the tree structure is faster than the pipelining ring structure for this particular application.>

Iterated function systems (IFSs) are capable of effectively describing complex shapes and textures by fractals. The interscale properties of such fractals are analyzed with the aid of the wavelet transform and general multiresolution analysis. The result is a formulation for homogeneous IFSs in general scale space which leads to a direct solution of the inverse problem of finding the IFS which best represents a given function. Multiscale techniques that are used in the analysis are discussed. Some previous results from the IFS literature are introduced.>

The bearing estimation problem is mapped onto the Liapunov energy function of the Hopfield model neural network. However, the Hopfield model implements a gradient descent algorithm, and, in common with all such algorithms, it is liable to find a local minimum rather than the desired global minimum. To overcome this problem three modifications, gain annealing, iterated descent, and stochastic networks, have been proposed. The modifications to the neural algorithm are outlined and simulated, and results are presented to show their convergence properties in the context of the bearing estimation problem.>

The maximum eigenvalue varied rate criteria (MEVRC) method is presented as a novel approach for determining the number of harmonics in a given signal. In contrast to the AIC and MDL criteria, MEVRC does not use a bias term. It is based on the eigenvalue varied rate of the smallest eigenvalues to estimate the number of signals. Simulation results indicate that the performance of MEVRC is better than AIC in normal signal-to-noise ratio (SNR) and sample situations, and better than the MDL in adverse conditions.>

A fast implementation of the ESPRIT algorithm, incorporating parallel processing, is described. ESPRIT is a novel, subspace fitting, parameter estimation algorithm which is used for obtaining high-resolution, unbiased estimates of the frequencies and powers of complex sinusoids in noise. The algorithm involves complex decompositions which require extensive computation. In order to use the algorithm in close to real-time situations, it is altered to facilitate faster computation and the use of a parallel architecture. The altered algorithm is decomposed onto an array of Inmos T800 transputers and its performance compared to that of the standard TLS-ESPRIT through simulations.>

A system is being developed for automatic realignment of the patient in magnetic resonance imaging (MRI). Two images of the patient taken at different times are registered. Correction is made for any translation or rotation of the slices taken at a later date with respect to the master set (the scan taken on the first visit). Once the rotational movement is worked out, the slices are rotated through this angle and the registration parameters computed. This has motivated the development of a new technique for the rotation of MR images that is fast and fairly accurate. MR scans of the head are littered with edges, so that rotation algorithms using interpolation techniques based on the bilinear insertion procedure have the effect of blurring the image. The proposed method uses an interpolation technique that preserves the edges and was found to be superior in accuracy and speed to the bilinear interpolation rotation method.>

For fast and numerically stable algorithms, array processors with floating point multiplication, division, and square rooting are necessary. The authors consider the use of the arithmetic operation of square rooting in the QR algorithm as used in many linear algebraic signal processing algorithms. Rather than reformulating the algorithms to be square root free with the inherent problems of numerical instability, loss of orthogonality, and overflow/underflow, the square root is reconsidered from first principles and arrays are designed that are as fast and have a smaller chip area than the analogous division arrays. This implies that implementations such as square foot free Givens rotations should not be considered in an application-specific integrated circuit or similar design due to their potential instability and susceptibility to overflow.>

A novel approach for the inspection of composite materials is presented, which utilises image processing techniques for the characterisation of flaws in composites. The technique is based on transient thermography coupled to the processing of infrared images obtained by exciting a composite specimen by a thermal pulse. The authors focus on the reconstruction algorithms developed for subsurface defect location, and overview the associated image processing techniques, as well as the experimental set-up for testing composites and generating thermal images.>

Two modifications to the neural-network algorithm originally proposed by J.J. Hopfield (1982), gain annealing and iterated descent, are proposed that yield better convergence to the global minimum. Simulation results are presented to illustrate the performance of the proposed algorithm for bearing estimation.>

A method of estimating the optical flow in a noisy image sequence is presented. In this approach a gradient matching procedure is effectively combined with an image differencing technique to obtain displacement vectors between two successive frames. The resulting displacement field is dense, and leads to compact depth maps. The computational cost is shown to be lower than in a simple correlation method. A maximum likelihood estimate is used to characterise the probability of match which can give a local estimate of the noise distribution in the frame sequence. Such knowledge can be used effectively for the determination of displacement vectors in subsequent frames of the image sequence.

A novel approach for real time inspection of composite materials is presented, which utilises image processing techniques for the characterisation of flaws in composites. The technique is based on transient thermography coupled to the processing of infrared images obtained by exciting the specimen by a thermal pulse. This paper discusses the image reconstruction theory used for defect location, associated image processing techniques, the numerical models for solving the propagation of heat through flawed composites as well as the experimental set-up for testing composites and generating thermal images.

This paper examines the computation of pixel level image processing algorithms on a linear array of processors. Occam algebra is used to show how convolution type image processing algorithms, which are fundamentally two dimensional in character can be computed on linear arrays. The paper concludes by considering the hardware/software tradeoffs that are possible for arrays implemented using transputer chips and compares the results with other parallel image processing architectures. The principle advantage of the proposed architecture is that it reduces the necessity for a frame store, and leads to very low values of data latency.

This paper describes a coding strategy which stores digital images in terms of their contours. The data in the original raw image is reorganised into a set of 1-d trajectories and these are normalised in length and approximated by low rank models. It is the coefficients of these models which are retained. The second order statistical information is used to derive a vector codebook containing standard sets of coefficients, and each contour is allocated the codeword of the nearest standard vector. In this way a codebook of generating functions for standard shapes is obtained. The problems of non closing contours are addressed along with interpolation methods to produce a reconstructed image from a contoured image.

This paper approaches the problem of designing dynamic constraints for use with an ill-conditioned image restoration process. A rule-based system is proposed that exploits feedback between image analysis and the low level analytic process. This allows the system to update the constraints by reasoning about their validity in the context of an image description generated from an a-priori class model and partial restorations of the image. The system has been used to show the effect of using structural constraints to define the solution space as an alternative to static, statistically-derived a-priori bounds on the solution.

An algorithm based on the gradient descent approach is proposed for adaptive signal processing. The algorithm outperforms the least mean square algorithm (LMS) in terms of convergence speed and misadjustment noise, and meets the performance characteristics of the conjugate gradient (OG) and recursive least squares (RLS) methods without the attendant computational complexity. It is simple to implement and lends itself to real time processing. The algorithm can be implemented using lattice or tap-delay line structures. In this paper extensive computational results are presented to illustrate the algorithm performance in terms of convergence properties, and its applications in adaptive noise cancelling and adaptive spectral analysis.

There is currently significant interest in the use of highly concurrent processor arrays to implement signal processing algorithms as they lead to a substantial reduction in the computational time. This paper shows how the OCCAM programming language can be used to aid the design of such architectures. This design approach has several interesting features: systolic and wavefront arrays can be designed, and since the method is OCCAM based, the algebra used to describe the computational structure is also the code required to program the algorithm on an array of INMOS IMS T424 transputers, INTEL 8086 processors, or Motorola 68000 processors.

This paper presents an attempt to combine statistical processing with symbolic processing for the detection of anomalous regions in images with noisy background. Symbolic inferences are used to provide expectations about the image to allow optimisation of statistical routines under these conditions. The approach employs a model of the image formation process and data about the domain.

The signal subspace algorithm, based on functions of the eigenvectors and eigenvalues of a data covariance matrix, is often used as a "high resolution" parameter estimator. In this paper, the resolving power of a signal subspace method is studied. By employing the statistical distributions of the eigenvectors of a sample covariance matrix, a measure of the expected resolving power of the MUSIC source direction estimator is obtained. The analysis shows that the ability of the MUSIC algorithm to resolve two closely spaced sources incident on an array of sensors is strongly linked to the observation time, the signal to noise ratio, and the separation between the sources.

Expressions are presented which quantify the effects of mis-estimation of the support of the object function on the regularised minimum-norm solution to the discrete bandlimited extrapolation problem. Computational results based upon these expressions are used to demonstrate the variation in the sensitivity of the regularised minimum-norm solution as a function of the regularisation parameter and space-bandwidth product.

Conventional processors for analysing pulsed Doppler returns employ either fixed digital filter schemes for target location, or FFT processors for identifying target velocities. In this paper a new configuration is proposed for processing radar returns which employ a lattice structure. Based upon the underlying probability distribution of the complex reflection coefficients; closed form analytical expressions, maximising detection probabilities for given false alarms, are presented which allow the use of a lattice as a detector by setting thresholds on the associated reflection coefficients in order to identify a signal from background noise. Using clutter models for ground clutter and weather clutter it is shown that in some circumstances the lattice can be used to identify a target in clutter environment.

A new design method for 2D circularly symmetric FIR filters is presented which is based upon the constrained optimisation of the passband volume of the filters magnitude squared function. It is shown that if a quadratic constraint is employed the resulting filter coefficients are related to a set of 2D prolate spheroidal sequences. These filters have the property of concentrating the maximum energy into the filter passband and have a phase characteristic which is linear. The paper also extends the theory of discrete prolate spheroidal sequences to two dimensions.

A recurrence relation is used to model image blurring, in place of the more conventional convolution. It is shown how this choice can lead to an efficient image restoration filter, by taking advantage of the recursive formulation.

This paper addresses the problem of removing time varying multiples from reflection seismic data. An adaptive recursive lattice algorithm is proposed for adaptively deconvolving the multiple events from the data. The technique involves the use of a finite memory filter which propagates through the time series. This memory window is set up such that both the future and past neighbouring statistics are used for computing the reflection coefficients. A comparison is made between conventional deconvolution techniques with synthetic and field reflection data. To achieve 'D' step prediction filtering a two step filtering structure is proposed which over comes any orthogonality violation. Results on seismic data are included to illustrate the performance of the algorithm.

This paper considers some asymptotic statistical properties of covarianee eigenstructure spectral analysis techniques. It is shown that when the signal model is of the appropriate form, and the observations are Gaussian, the signal parameter estimates, obtained by locating the nulls in the eigen-spectrum, are asymptotically zero mean normal random variables. Based on this observation, the paper then considers the formation of confidence regions for the signal parameters. The paper presents the general case of a multi-dimensional eigenstructure algorithm, which estimates one or more parameters of each signal in the observed data.

Systolic architectures for signal processing are of great interest as they offer a considerable speed improvement over traditional Von-Nuemann computing architectures, and are particularly suitable for VLSI implementation due to ensuing simple and regular communication structures [1]. This paper presents an architecture for computing the Fast Fourier Transform (FFT) using a systolic processor which incorporates an elevator concept to circumvent the requirements for global communication inherent in conventional FFT implementations. The proposed algorithm is shown to be highly efficient in terms of both hardware and computation time. Architectures are further suggested for the real time computation of 2D functions such as the Wigner Distribution and the Complex Ambiguity function by using the systolic FFT processor coupled with an input characterising array.

Frequency domain techniques for pulsed Doppler radar data for target detection has received significant attention in the last few Years, and techniques employing FFT and Maximum Entropy methods (MEM) are much in favour. Some recent work [1,2] directed at comparing FFT techniques with MEM seem to indicate anomalies in the use of high resolution spectral estimation for target detection. Most of these techniques employ a blind decomposition of phase detected radar returns into frequency components, and make assessments based on computer simulations. This paper presents new results in the use of lattice processors for detecting signals in noise. A new approach to processing Doppler returns is proposed which involves the application of thresholds to the reflection coefficients in the lattice processor. A study of the underlying probability distribution of the reflection coefficients is included, which allows an optimal choice of threshold setting for a given false alarm rate, and probability of detection.

The paper presents two methods for the determination of 2D eigenfilter spectra, both of which can be viewed as a 2D extension of the conventional Pisarenko technique. The first approach taken is to formulate the problem as the design of a 2D moving average filter whose output energy must be minimised subject to a specified constraint. A second 2D eigenspectra technique can be developed by modelling 2D sinusoids in white noise. In both cases the underlying process spectra is determined from an eigenvector of an autocorrelation matrix. It is shown that when the second technique is used the autocorrelation matrix required can always be of minimal size.

In this paper an adaptive lattice structured filter for processing of narrowband spatial data as received by an array of sensors is described, This spatial lattice filter is shown to take on a triangular structure in the spatial domain, composed of N(N-1)/2 individual two multiplier lattice cells for an array of N sensors. This is considerably more than an equivalent order lattice for the prediction of a statioary time series, and is neccesary to cope with the possibly non-Toeplitz covariance matrices that arise from multipath signals and irregularly spaced arrays. It is shown however that in a hardware implementation of the filter, advantage can be taken of parallel procesing, reducing the total computational cost to O(N) processing cycles. It is also shown how the same lattice structure can be used to evaluate the prediction coefficients for the sensor signals from the lattice reflection coefficients - this being efficiently carried out with the use of systolic processing.

The problem of deconvolving severely smeared images is considered in this paper. It is shown that classical techniques for deconvolution are, for some data sets, inadequate due to the severity of the smearing. The reasons for this are investigated and a measure is obtained for the signal to noise ratio improvement of estimators operating on isotropic smearing on a circular domain. The issue of what can be estimated from such data sets is then investigated and an adaption of the Kormylo-Mendel single most likely replacement algorithm is proposed as a method of estimating sparse sources from such data sets.

This paper details an analysis for the presence of spectral windows associated with the Maximum Entropy (MESA) Spectral Estimator. By embedding the derivation of the spectral estimator into a constraint optimisation framework a relationship is established between the true spectrum of a second order stationary stochastic sequence and the Maximum Entropy spectrum. This relationship is shown to lead to a spectral window which appears in the reciprocal of the MESA estimator. The structure of the window is based upon a Dirichlet type kernel and depends upon the prediction filter length for a large class of data sequences. Computed results are included which verify the presence of these spectral windows.

A new method is presented for resolving the directions of multiple sources whose emitted radiation is received by a linear array of sensors. A linear predictive type ARMA model is derived for the received signals where the data represents incident radiation from multiple sources plus measurement noise. The model facilitates identification of coherent sources from non-equispaced sensor data which more realistically represents the all important practical radar situation, and also accounts for multipath effects.

This paper extends the constrained optimisation image reconstruction techniques to curved-ray projections for a number of scanning systems and for 2-D and 3-D reconstructions. The work presented is general in that it includes the previous results on parallel-ray and divergent-ray geometries as special cases. Treating the problem in Hilbert space a general cost function is optimised and the solution is applied to a number of practical criteria. This has led to a convergent iterative algorithm which evaluates the associated Lagrange multiplier functions and establishes the reconstruction for all scanning geometries. Further a generalised Projection Slice Theorem is introduced based on a new Projection Transform.

This paper describes the development of a multi-path and multimode propagation model for NDE. The mode content of a multipath is estimated using Maximum Likelihood (ML) estimation based on a receiving array delay vector. The model analysis allows a defect to be located by employing both the mode estimates and a ML estimation of the shear and longitudinal propagation angles. Some aspects of the attendant processing hardware are also discussed.

This paper examines constrained optimization designs of infinite impulse response (I.I.R.) digital filters and introduces two new design methods, including one based upon discrete prolate spheroidal wave functions. The designs employ a recently introduced model for the magnitude squared function of an all pole filter, the coefficients of which can be obtained using optimization techniques subject to either linear or quadratic constraints. The generalized design methods introduced in this paper allow the design of low pass and high pass filters and since the filters can be implemented as lattice filters they should have low finite word length errors. The design methods introduced for 1 Dimensional filters can be extended to 2 Dimensions and preliminary results are included.

This paper is concerned with the techniques required for the acquisition and processing of signals arising in ultrasonic non-destructive testing of steel structures in the underwater environment. The primary aim is to minimize diver participation in the inspection and interpretation of results. The resulting systems, which involve significant use of microprocessor hardware were designed to operate with equal facility with either single or multiple ultrasonic channels, the latter being important with the ancillary requirements of beam steering and beam shaping. The paper describes two approaches to the detection and location of faults, one concerned with the use of large ultrasonic crystals has led to the development of the Strathclyde SHOE, the other employs ultrasonic arrays for area scanning via beam steering. An analysis is included for signal returns on the arrays for estimating the location of a reflection point which leads to enhanced range resolution.

This paper is concerned with problems of enhancing the look-direction signal in the presence of spatially distributed interference sources and sensor noise, when a multi channel processor is employed subject to the constraint that it has a desired frequency response for look-direction signals. The multi-channel algorithms to be presented possess stage by stage decoupling and do not involve an arbitrary size of the step length, unlike conventional tapped delay line algorithms. General implementation of constraints are discussed in terms of multi channel adaptive lattice algorithms. Different structures are proposed for the constraint algorithms based, on a direct form and a normalised form, which may be applied to broad-band sensor data. Single channel multi-sensor adaptive lattices are considered for narrow-band array data processing and their performance compared with eigenfilter (Pisarenko) type processors.

A new model for piezoelectric ultrasonic transducers is proposed. It is derived from the fundamental piezoelectric equations and it has the following features: i) It is valid over a wide range of frequencies. ii) It is applicable in both transmission and reception modes. iii) It involves realizable elements which are readily simulated. iv) It involves feedback mechanisms which clearly relate pressure and voltage interactions. The model has been widely investigated using computer simulation, water tank measurements and photo -elastic visualisation studies using glass models. Some of the results of such investigations are presented.

Laser Doppler velocimeters provide a noncontact method for the measurement of velocities. These measurements are, in general, corrupted by noise arising in the transmission and reception of the optical carrier. An analysis is presented for the effect of noise on the output of the velocimeter system under conditions of high scattering center concentrations and Steady flows. Principally, the so-called Gaussian beam optical setup is considered where the intensity distribution of the intersecting laser beams at the observation volume (fringe pattern) is Gaussian. Output noise spectral densities, noise power, and signal-signal-to-noise ratios are derived in terms of the input signal (Doppler carrier)-tonoise ratios and bandwidths of the processing systems. Also, threshold levels for input signal-to-noise ratios are determined and the photodetector performance analyzed. Finally, analogies are drawn where the current analysis is applicable to other areas of communication.