Pub Date : 1995-05-21DOI: 10.1364/optcomp.1995.owa1
J. Heanue, M. C. Bashaw, L. Hesselink
We discuss the experimental performance of a digital holographic data storage device and architectural and materials issues related to achieving large capacity and low bit error rates.
我们讨论了数字全息数据存储设备的实验性能，以及与实现大容量和低误码率相关的架构和材料问题。
{"title":"Volume Holographic Storage and Retrieval of Digital Information","authors":"J. Heanue, M. C. Bashaw, L. Hesselink","doi":"10.1364/optcomp.1995.owa1","DOIUrl":"https://doi.org/10.1364/optcomp.1995.owa1","url":null,"abstract":"We discuss the experimental performance of a digital holographic data storage device and architectural and materials issues related to achieving large capacity and low bit error rates.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-05-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116991416","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We are developing a class of optical phased-array-radar processors which use the large number of degrees-of-freedom (DOF) available in three-dimensional photorefractive volume holograms to time integrate the adaptive weights in order to perform beam-steering and jammer-cancellation signal-processing tasks for very large phased-array antennas[1,2]. For a large broadband phased-array antenna containing 1000s of array elements, beam steering and jammer cancellation in a dynamic signal environment represents an extremely demanding signal processing task well beyond the capabilities of microelectronic digital signal processing because of the large number of DOF required for adaptation. The three-dimensional nature of the signal environment (2 angle-of-arrival and frequency) represents a signal processing problem which maps well into a highly parallel optical processing architecture utilizing photorefractive volume holograms. The beam-steering and jammer-nulling processor we present uses relatively simple components; two photorefractive crystals, two single-channel high-speed detectors, and two single channel acousto-optic Bragg cells. The bandwidth capabilities of these components approach a GHz allowing the processing of wide-band signals. The required number of processor components used for implementing the adaptive algorithm is independent of the number of elements in the phased-array in contrast to traditional electronic or acousto-optic approaches[4,5], in which the hardware complexity of the processor scales in proportion to array size. We describe the two main subsystems of the processor, the beam-forming and the jammer-nulling subsystems, and present results demonstrating simultaneous main beam formation and jammer suppression in the combined processor.
{"title":"Adaptive Beam-Steering and Jammer-Nulling Photorefractive Phased-Array Radar Processor","authors":"A. Sarto, R. Weverka, K. Wagner","doi":"10.1117/12.177419","DOIUrl":"https://doi.org/10.1117/12.177419","url":null,"abstract":"We are developing a class of optical phased-array-radar processors which use the large number of degrees-of-freedom (DOF) available in three-dimensional photorefractive volume holograms to time integrate the adaptive weights in order to perform beam-steering and jammer-cancellation signal-processing tasks for very large phased-array antennas[1,2]. For a large broadband phased-array antenna containing 1000s of array elements, beam steering and jammer cancellation in a dynamic signal environment represents an extremely demanding signal processing task well beyond the capabilities of microelectronic digital signal processing because of the large number of DOF required for adaptation. The three-dimensional nature of the signal environment (2 angle-of-arrival and frequency) represents a signal processing problem which maps well into a highly parallel optical processing architecture utilizing photorefractive volume holograms. The beam-steering and jammer-nulling processor we present uses relatively simple components; two photorefractive crystals, two single-channel high-speed detectors, and two single channel acousto-optic Bragg cells. The bandwidth capabilities of these components approach a GHz allowing the processing of wide-band signals. The required number of processor components used for implementing the adaptive algorithm is independent of the number of elements in the phased-array in contrast to traditional electronic or acousto-optic approaches[4,5], in which the hardware complexity of the processor scales in proportion to array size. We describe the two main subsystems of the processor, the beam-forming and the jammer-nulling subsystems, and present results demonstrating simultaneous main beam formation and jammer suppression in the combined processor.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133105510","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Optical memory disks have been developed in recent years as mass storage media for audio, video, and computer memory applications. Write-once systems are already widely used, and reprogrammable systems are now starting to become commercially available as well. In all the existing systems the information stored in the optical disk is recorded and readout serially by focusing a laser beam on a single pixel. With an optical memory however it is possible to illuminate the disk with an extended beam and readout (as well as record in principle) large amounts of data in parallel [1]. This distinction between serial and Parallel Readout Optical Disks (PROD) is schematically shown in Fig.1. If the potential of PRODs is realized in practice it can eliminate the bottleneck that currently exists between mass memory and the information processing portion of a computer and thus greatly impact the speed with which computers can execute memory intensive problems. There are three main issues that we will address in this paper: The suitability of commercially available disks for this applications including the experimental characterization of a prototype magnetooptic system from SONY, the limitations imposed on parallel access due to the optical system, and the types of problems and computer architectures that can make effective use of the PROD capability.
{"title":"Parallel Readout of Optical Disks","authors":"D. Psaltis","doi":"10.21236/ada256625","DOIUrl":"https://doi.org/10.21236/ada256625","url":null,"abstract":"Optical memory disks have been developed in recent years as mass storage media for audio, video, and computer memory applications. Write-once systems are already widely used, and reprogrammable systems are now starting to become commercially available as well. In all the existing systems the information stored in the optical disk is recorded and readout serially by focusing a laser beam on a single pixel. With an optical memory however it is possible to illuminate the disk with an extended beam and readout (as well as record in principle) large amounts of data in parallel [1]. This distinction between serial and Parallel Readout Optical Disks (PROD) is schematically shown in Fig.1. If the potential of PRODs is realized in practice it can eliminate the bottleneck that currently exists between mass memory and the information processing portion of a computer and thus greatly impact the speed with which computers can execute memory intensive problems. There are three main issues that we will address in this paper: The suitability of commercially available disks for this applications including the experimental characterization of a prototype magnetooptic system from SONY, the limitations imposed on parallel access due to the optical system, and the types of problems and computer architectures that can make effective use of the PROD capability.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"123 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131659279","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-05-22DOI: 10.1364/optcomp.1991.me23
M. Taghizadeh, J. Turunen, B. Robertson
Recent progress in designing and manufacturing space-invariant optical array generators is described. We begin by demonstrating Dammann gratings [1] that generate even-numbered arrays as large as 128x128, and odd-numbered arrays of up to 201x201 spots. The concept of a hybrid hologram [2] is applied to the fabrication of array generators, and extremely high-efficiency (close to 90%) components are obtained. Several novel types of array generators with multiple phase levels are introduced. These can e.g. reconstruct arrays with different fan-out at different angles of incidence. The application of rigorous diffraction theory to design highly efficient and compact array generators is also discussed.
{"title":"Passive Optical Array Generators","authors":"M. Taghizadeh, J. Turunen, B. Robertson","doi":"10.1364/optcomp.1991.me23","DOIUrl":"https://doi.org/10.1364/optcomp.1991.me23","url":null,"abstract":"Recent progress in designing and manufacturing space-invariant optical array generators is described. We begin by demonstrating Dammann gratings [1] that generate even-numbered arrays as large as 128x128, and odd-numbered arrays of up to 201x201 spots. The concept of a hybrid hologram [2] is applied to the fabrication of array generators, and extremely high-efficiency (close to 90%) components are obtained. Several novel types of array generators with multiple phase levels are introduced. These can e.g. reconstruct arrays with different fan-out at different angles of incidence. The application of rigorous diffraction theory to design highly efficient and compact array generators is also discussed.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122221412","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-05-22DOI: 10.1364/optcomp.1991.wa3
J. Wullert, Eung G. Pack, J. S. Patel
Neural networks, characterized as a large number of highly interconnected simple processors, can be trained by varying the strength (weight) of the interconnections (synapses) between the simple processors (neurons). Several holographic optical systems have physically demonstrated this capability previously.[1][2][3][4] Since neural networks are trained by example rather than programmed with specific rules, they are likely to be able to generalize, or recognize patterns that do not exactly match those used for training. Such generalization is important in real world pattern- recognition problems where the size, orientation, position and background cannot be determined in advance.
{"title":"Generalization in an Optical On-Line Learning Machine","authors":"J. Wullert, Eung G. Pack, J. S. Patel","doi":"10.1364/optcomp.1991.wa3","DOIUrl":"https://doi.org/10.1364/optcomp.1991.wa3","url":null,"abstract":"Neural networks, characterized as a large number of highly interconnected simple processors, can be trained by varying the strength (weight) of the interconnections (synapses) between the simple processors (neurons). Several holographic optical systems have physically demonstrated this capability previously.[1][2][3][4] Since neural networks are trained by example rather than programmed with specific rules, they are likely to be able to generalize, or recognize patterns that do not exactly match those used for training. Such generalization is important in real world pattern- recognition problems where the size, orientation, position and background cannot be determined in advance.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"1260 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132603303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-05-22DOI: 10.1364/optcomp.1991.tuc3
D. Chiarulli, S. Levitan, R. Melhem
This experiment is based on two properties of optical signals, unidirectional propagation and predicatable path delay. Using these properties, logic systems can be devised in which information is encoded as the relative timing of two optical signals. Coincident pulse addressing is an example of such a system. In this case, the address of a detector is encoded as the delay between two optical pulses which traverse independent optical paths to a detector. The delay is encoded to correspond exactly to the difference between the two optical path lengths. Thus, pulse coincidence, a single pulse with power equal to the sum of the two addressing pulses, is seen at the selected detector site. Other detectors along the two optical paths for which the delay did not equal the difference in path length, see both pulses independently, separated in time.
{"title":"Demonstration of an All Optical Addressing Circuit","authors":"D. Chiarulli, S. Levitan, R. Melhem","doi":"10.1364/optcomp.1991.tuc3","DOIUrl":"https://doi.org/10.1364/optcomp.1991.tuc3","url":null,"abstract":"This experiment is based on two properties of optical signals, unidirectional propagation and predicatable path delay. Using these properties, logic systems can be devised in which information is encoded as the relative timing of two optical signals. Coincident pulse addressing is an example of such a system. In this case, the address of a detector is encoded as the delay between two optical pulses which traverse independent optical paths to a detector. The delay is encoded to correspond exactly to the difference between the two optical path lengths. Thus, pulse coincidence, a single pulse with power equal to the sum of the two addressing pulses, is seen at the selected detector site. Other detectors along the two optical paths for which the delay did not equal the difference in path length, see both pulses independently, separated in time.","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114885763","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1992-05-22DOI: 10.1364/optcomp.1991.me7
M. Robinson, K. Johnson, D. Jared, D. Doroski, S. Wichart
Optical implementations of one-layer, perceptron-like neural networks have been shown to be very successful at associating pattern/target sets despite large system errors [1,2]. It has also been shown that large systems can be realized with such architectures (≥4 x 104 interconnections [2,3]), and appreciable processing speeds have been demonstrated (>108 interconnections/sec [4]). However, single layer networks are limited due to their inability to associate patterns that are not linearly separable. A more general network is the two layer network, which is able to model arbitrary functions, and create any decision boundary within the input vector pattern space [5]. In order to implement such a network, it is necessary to perform a nonlinearity at the hidden layer before performing a subsequent matrix multiplication. In general, optical materials performing fast nonlinear processing require high optical powers. Hybrid opto-electronic devices can perform nonlinear operations at moderate speeds and low optical powers [6].
尽管存在较大的系统误差，但单层感知器类神经网络的光学实现已被证明在关联模式/目标集方面非常成功[1,2]。研究还表明，使用这种架构可以实现大型系统(≥4 x 104互连[2,3])，并且已经证明了可观的处理速度(>108互连/秒[4])。然而，单层网络由于无法将非线性可分的模式关联起来而受到限制。更一般的网络是两层网络，它能够对任意函数建模，并在输入向量模式空间内创建任何决策边界[5]。为了实现这样的网络，在执行后续的矩阵乘法之前，有必要在隐藏层执行非线性。一般来说，进行快速非线性处理的光学材料需要较高的光功率。混合光电器件可以在中速、低光功率下进行非线性运算[6]。
{"title":"Custom designed electro-optic components for optically implemented, multi-layer neural networks","authors":"M. Robinson, K. Johnson, D. Jared, D. Doroski, S. Wichart","doi":"10.1364/optcomp.1991.me7","DOIUrl":"https://doi.org/10.1364/optcomp.1991.me7","url":null,"abstract":"Optical implementations of one-layer, perceptron-like neural networks have been shown to be very successful at associating pattern/target sets despite large system errors [1,2]. It has also been shown that large systems can be realized with such architectures (≥4 x 104 interconnections [2,3]), and appreciable processing speeds have been demonstrated (>108 interconnections/sec [4]). However, single layer networks are limited due to their inability to associate patterns that are not linearly separable. A more general network is the two layer network, which is able to model arbitrary functions, and create any decision boundary within the input vector pattern space [5]. In order to implement such a network, it is necessary to perform a nonlinearity at the hidden layer before performing a subsequent matrix multiplication. In general, optical materials performing fast nonlinear processing require high optical powers. Hybrid opto-electronic devices can perform nonlinear operations at moderate speeds and low optical powers [6].","PeriodicalId":302010,"journal":{"name":"Optical Computing","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116875743","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：