I am happy to be here this morning with this distinguished technical audience of the Society of Photo-Optical Instrumentation Engineers. You have recognized a very important problem area and I think that through this conference or seminar, you will address many technical aspects of the problems associated with Command, Control, Communications and Intelligence (C³I). C³I is a term that is gaining more and more visibility and acceptance in all areas of military operations. In the past, the transfer of technology into military operations has been concentrated on force effectiveness. Our leaders now recognize more fully that force effectiveness depends to a very high degree on the command and control functions which in turn need to know the status of enemy as well as friendly forces. Perhaps the best way to start this discussion of C³I is to arrive at a definition. Unfortunately, I'm sure that if I ask each of you to give me a definition of C³I, I would have just as many definitions as there are people in the audience. The situation is similar to the old story about the three blind men trying to describe an elephant. I would like to paraphrase a few remarks of Julian Lake in a recent editorial in Military Electronics/Countermeasures Magazine. As he points out, C³ is many things to many people. To the intelligence specialist who is so wrapped up in his intelligence community activity, C³I is simply an extension of the modern applications of intelligence. In fact, the intelligence officer feels that he is the actual center of the C³I structure. On the other hand, the communications specialist thinks that communications is the actual heartbeat of C³I activity, and this is right to a point because communications is a fundamental building block of the C³ function. The computer specialist will point out that C³I is merely a product of the computer age. One reason there was very little done about C³I in the past was the nonavailability of computer techniques and consequently this is what has really made C³I possible. The radar man believes that this other stuff is absolutely useless without information about the activities of the enemy other than those derived through the intelligence community. The display technologist feels that without some means of presenting this information to the commander, the process is meaningless. The commander needs to absorb a large amount of information quickly. Jerry Lawson of the Naval Electronics Systems Command has pointed out that in a typical situation, there will be at least twenty to one hundred objects about which the commander must be informed. For each of these objects, it can be shown that he will need an absolute minimum of about 300 bits of data before he can be an effective transducer of information into decisions and directives. Studies have shown however, that the human brain can process only 25 - 40 bits of information per second when they are presented as a time sequential string such as in a printed message. This implies that it would take about 15 minutes for the commander to understand what was going on; and the picture is changing continuously. Yet, our common experience tells us that we can easily process three million bits per second when they are presented as a two-dimensional picture. Thus, we have no alternative but to present the commander with a picture, a geographic display, so that he can absorb the large number of spatial relationships among the things with which he must deal. Moreover, on a plot of this type, we can express much of the data by the size and the color of the symbols we use to mark the location of things. Consequently, three important measures of the efficiency of this process are the locational accuracy, the informational accuracy and the timeliness of the picture which is presented.

A brief presentation of the classic multichannel frequency plane, joint transform and hybrid optical signal processors is followed by a description of the classic folded spectrum analyzer and Spann correlator. A general review of advanced acousto-optic signal processors using the chirp-Z transform, and a description of the triple product processor is given. Examples, features and C³I applications of the various time, space and hybrid time and space integrating systems are provided.

Political and military imperatives will continue to push the technology for faster processors, smarter information systems, larger memories, and greater communications capacity to ensure the reliable and secure control of our military forces. The advent of satellite surveillance and reconnaissance and the continual development of more sophisticated radar systems has opened the way to providing military commanders the capability of real-time monitoring of military conflicts. Today's research must push an emerging technology capable of transmitting and processing vast quantities of information and displaying the most pertinent command information at a centralized command center. One such technology is optical processing, or better yet, hybrid optical-analog/electronic-digital processing. Optical systems offer parallel computations with speeds unattainable with present digital computer technology, and the hybrid systems combine this speed and data handling capability with the computational flexibility and accuracy of digital processing. The Air Force sponsored research program is aimed toward increasing the flexibility of optical processing through the support of scientific endeavors in such areas as non-linear optics, optical feedback, space variant processing, and optical computations. The research spans a broad realm of efforts from feasibility studies all the way to the fabrication of novel devices. This paper will describe the currently active efforts in this program.

During the past two decades, a number of methods have been proposed for performing bipolar spatial filtering operations with incoherent optical systems. These methods are conveniently categorized as to being based on geometrical optics or on diffraction optics. Some methods require a mechanical scanning of the input object, others are fully parallel in their operation. Still others are optically parallel, but employ an image plane scan to provide an electrical signal representation of the bipolar processed image. This paper reviews these methods and, to a limited extent, compares them.

It is often desired to transmit video signals over a data channel. However, when a standard 4 MHz bandlimited video signal is sampled at the Nyquist rate and encoded using four to six quantization levels (PCM) the data rate becomes 48-64 Mbs. This high digital bandwidth is usually unacceptable for most applications. Various digital alternatives such as delta modulation have been used to lower the data rates to about 22 Mbs. Computer simulations have shown that transform encoding can reduce this digital rate further. Additional reduction in the rate is usually desired to allow the use of available analog bandlimited video channels. In this paper we report on the effects of analog two dimensional spatial filtering in conjunction with delta modulation (DM) on digital bandwidth compression. Experimental and computer simulated results of hybrid DM enhancement encoder is presented. Results of computer simulation of a hybrid cosine transform encoder we also presented.

Basic principals of radar systems are introduced and developed for treatment in adaptive systems. Parameters which must be weighted in order to maximize signal-to-noise ratio are described. These parameters include aperture illumination, instantaneous bandwidth and pulse repetition rate. They are related to antenna field pattern, range resolution, and doppler resolution. The treatment shows how these parameters are weighted in a changing noise environment for optimum system performance. Algorithms for generating these weights are illustrated in vector and matrix notation. This discrete set of weights is extended to accommodate the large number element case. Here, a continuum set is approached and the adaptive equations assume an integral form. The adaptive algorithm is thus presented over a wider version, i.e. matrix or integral notation, for consideration by the optical processing community.

A general presentation of the holographic processing of antenna data is followed by a discussion of the practical possibilities using a 2-D Pockels modulator as real-time input. The complete processing of an active sonar circular array is described and other applications are suggested for passive sonar and active radar antennas, with consideration of actual technology limitations.

A coherent optical correlator (operating on the heterodyned received signals from a phased array and a non-coherent iterative vector-matrix electro-optical processor (operating on the covariance matrix of received array signals) are described as candidate advanced processors for adaptive phased array radar.

Spread-sprectrum techniques are widely used in state-of-the-art communication systems for suppression of interference. In this paper we will establish the characteristics required for signal-processing devices used in such systems. Typical spread-spectrum waveforms will be described to emphasize those characteristics which lead to good system performance. The synchronization problem will be reviewed to show that this is often the most difficult aspect of a communication system design and strongly influences the choice of devices. The performance of any device in a system will reflect some features of the device which are undesirable. Spurious responses can cause errors in detection or demodulation circuitry. A limited dynamic range for a signal-processing device might have the effect of lowering the apparent interference rejection of the system.

Present-day processing requirements for radar and communication signals range from real-time, large bandwidth to large processing gain, all in a dense multisignal environment. By combining desirable features of acoustics and optics and in some cases charge-coupled devices (CCD), acousto-optic technology offers great promise of fulfilling these diverse requirements. This paper overviews some of the uses of acousto-optic technology, especially those using surface acoustic wave (SAW) devices in the areas of radar and communication signal processing. Two areas are covered, (1) Fourier transformation, including spectrum analysis, and (2) correlation.

Finite digital convolution (FDC) appears in the implementation of finite impulse response digital filtering, in auto-aid cross-correlation, polynominal multiplication and the multiplication of very large numbers. While there are several methods to implement FDC, when the lengths of the sequences to be convolved is a highly composite number, the discrete fast Fourier transform (FFT) approach is used. This approach requires generating, storing and truncating a large number of complex exponentials. Recent interest has centered on finding real basis numbers that preserve the properties of the FFT. By working in a finite field of integers with arithmetic modulo an integer M, a large class of new transforms, called number theoretic transforms, can be generated. These transforms are useful in applications where integer arithmetic is already being considered, such as spread-spectrum encoding, digital error-correction and data encryption, or where the data is digitally encoded in a finite number of bits. In this paper, residue arithmetic based number theoretic transforms will be considered.

This paper will serve to bring on board potential contributors to the task of developing real-time processing techniques for aircraft identification of friend, foe or neutral (IFFN) in a noncooperative target recognition (NCTR) environment. An overview of current digital processing techniques and data correlation for NCTR is presented. Reference is made to Figure 1, block diagram of Wideband Data Image Processor of which the first item of this paper will address. These items include: (1) Background of Wideband Radar, Digital Imaging Analysis , (1A) Motion Compensation, (1B) Image Generation, Scaling, (1C) Translation Identification, (2) Motivations and Advantages of Optical Processing for IFFN/NCTR, (3) RADC Development Efforts, (4) Current Deficiencies and Future Development Plans.

The coherent light valve is a video-input spatial light modulator developed for real-time optical processing. It was developed as a modified form of the commercial GE light valve TV projection system. The video input signal phase modulates the optical aperture by modulating a raster-scanned electron beam. At compatible TV rates (525/30) the CLV system has a bandwidth of 10-15 MHz and a dynamic range of 40 dB or more . Higher scan rates have demonstrated bandwidths up to 100 MHz, and a special CLV tube design can operate above 100 MHz. Recently compact rack-mounted processor configurations, suitable for user environments, have been developed. This paper reviews past and present work and offers some predictions regarding future developments of the CLV system.

This paper describes the integrated-optic implementation of a Bragg spectrum analyzer that employs the interaction between a coherent optical guided wave and a surface acoustic wave to determine the power spectral density of the input. The integrated-optic spectrum analyzer consists of an injection laser diode, a thin-film optical waveguide, waveguide lenses, a surface-acoustic-wave transducer, and a linear detector array with CCD readout. Design principles are given for selecting component parameters such as optical beam width, detector cell size, lens aperture and focal length, and acoustic transducer design so as to obtain specific rf resolution, spurious level, and signal-to-noise ratio. Design parameters are presented for a 750- to 1250-MHz spectrum analyzer with a resolution of 4 MHz and a 40-dB dynamic range. Also described in the paper is the development of state-of-the-art component technology for the spectrum analyzer.

This paper presents a review of recent progress on planar guided-wave acoustooptics with application to wideband real-time signal processing and communications. Key parameters of an acoustooptic Bragg modulator/ deflector relevant to signal processing and communications are discussed first. A projection of the performance figures obtainable with these parameters using the existing technology is also made. Implementation of the acoustooptic time-integrating correlators using integrated optics techniques is then described.

An analogue estimator is defined, on the basis of the Parzen window method. It performs estimates of the distribution function or probability density function, from a sample analyzed with correlation techniques. A simple optical implementation is derived, for the statistical analysis of one random parameter in 2-D scenes. Applications are suggested in pattern classification and image processing.

In several recent papers we have described a new kind of signal processing pattern recognition filter designed to be used exactly as matched filters are presently used. The new filter is usually superior to the matched filter in terms of both between-class discrimination and within-class tolerance. Because it has these properties and reduces to the matched filter for simple enough cases, the new filter is called the generalized matched filter (GMF). In this paper we give further detail on GMF derivation and operation. Particular stress is laid on the effects of the difference in optimization criteria for those simple cases for which both the matched filter and the GMF are defined.

Developers of surface-wave acoustooptic (AO) and acoustoelectric (AE) devices are currently competing with one another by conducting feasibility demonstrations which suggest substantial analog signal-processing capabilities. The development of one AE surface-acoustic-wave device, a wideband convolver, is currently in a more advanced state in terms of system-worthiness than are AO and other AE devices. However, in terms of future potential in most systems applications, it has not been apparent whether either class of devices has had a clear advantage. In this paper, figures of merit as well as design tradeoffs in the two technologies are compared. These comparisons focus on those areas of the competing technologies which will be most stressed as the devices become system-worthy components in the near future. Projections are made for the realization of practical devices capable of providing, with adequate dynamic range, the signal-processing gains suggested by the feasibility demonstrations. It is concluded that each system application must be considered individually to determine whether AE or AO devices are the most advantageous.

Improved noise performance, relaxed requirements on input devices, and increased dynamic range all follow from the use of noncoherent rather than coherent light for optical processing. Noncoherent light behaves differently than coherent light, and so different optical processing techniques are required. The emerging capabilities of such techniques includes matrix-vector multiplication, correlation, linear filtering and analog to digital conversion. Recent advances in the use of the noncoherent optical transfer function for linear filtering are discussed, and applied to the processing of raster formatted radar and communication signals.

Adaptive phase compensation receivers have been proposed for improving the performance of line-of-sight atmospheric optical communication systems. In this Paper we examine the system requirements dictated by the coherence parameters of low-visibility wavefronts, and describe a class of all-optical (essentially wireless) systems that are being investigated for high resolution, real time, wavefront phase compensation. Preliminary results from a simpler, single-element, opto-electronic test system that demonstrates the feasibility of the all-optical concept, are also reported.

This paper describes a new class of solid state devices for analog signal processing applications. These devices are achieved b integrating a surface acoustic wave (SAW) device with a suitable charge-coupled device (CCD) to make use of the advantages and compensate for the limitations of each. By combining the wide bandwidth of the SAW. with the long and flexible time handling capability of the CCD, it becomes possible to implement a family of unique devices capable of performing s as signal processing functions, including buffer memory, correlation, and matched filtering. We have fabricated and tested prototypes t perform. each. of these different functions. The fast-in, slow-out buffer memory has a 40-MHz bandwidth, an input signal duration of 3.5 os and an output clock rate of 100 kHz. The accumulating correlFltor has achieved a signal processing gain of 30 dB at a bandwidth of 20 MHz when or pseudohoise waveforms of 100 os duration. The CCD-programmable matched filter is capable of processing signals with a bandwidth of 40 MHz and a duration of 3.5 os. In each case the CCD signals a compatible with high-density, low-cost integrated circuits for handling the low-speed control and output aspects of the signal processing functions.

In recent years, considerable progress has been made in the development of components for wideband optical communication, using glass fibers as the transmission medium. The most dramatic advance has been in the loss in the fibers themselves, which has dropped from a minimum of about 1000 dB/km prior to 1970 to 0.2 dB/km today. Similarly, research in the field of integrated optics has led to the development of miniaturized, single-mode components for use with laser sources. Generation, modulation, directional coupling, switching, and detection of light are functions which have already been demonstrated with integrated optics devices. Improvements in bandwidth, switching and multiplexing capability, size, and reliability for fiber communications systems could result from these efforts.

Deep space surveillance is undergoing evolutionary change. The impetus for change stems from the dynamic and advancing nature of the threat which places severe demands on existing technology, and forces the development of new technology. This need to counter a higher order threat has stimulated the search for new technology developments that will provide greater performance. The emergence of a charge coupled device (CCD) technology offers great promise in the area of deep-space surveillance. To pursue further the capability of CCD's and to quantify their performance advantages, DARPA sponsored the Teal Amber I program in early 1976. Teal Amber I is designed to develop visible CCD technology and to conduct a CCD focal plane development and demonstration program. Phase II of the program, the field demonstration of a partially populated focal plane, has recently been completed and provides the impetus for this paper. As background, a description of the sensor system concept and the sensor performance goals are given. The focal plane signal processing used for background scene supression and target detection is described. The paper concludes with a description of the Phase II field demonstration and the performance results that were achieved.