1、光学矩阵器件Research Status and Hotspots of Optical Matrix Calculation DevicesAbstracts:When we say about matrix,we are mostly talking about multipliers. Vector matrix multipliers (VMMs) can be used to perform Fouriertransform and solve linear equations. They have wide applications in imageprocessing, bea
2、m forming, radar detection, wireless communication, etc.With the fast development of micro-electronic technologies during the pastdecades, the VMM computation speed has been improved significantly.However the continuous improvement in performances of CMOS chipsencounters bottlenecks in both physical
3、 size and manufacturing process. Aswell known, photonics has advantages of broadband, high speed, and lowpower consumption. It is easier to realize parallel processing. Hence,photonic technologies may provide a solution to solve the bottlenecks.Nowadays, researches on optical vector matrix multiplie
4、rs (OVMMs) havebecome a hot topic and attracted more and more attention in recent years.Because the multipliers are the most important devices of Optical Matrix Calculation Devices,Our focus is mainly on OVMMs.We will introduce the theory of matrix multiplication and we cansee some hotpots of Optica
5、l Matrix Calculation Devices.Introduction:Traditional electronic computing technology showsmore and more shortcomings with the increase ofrequirements for the computing capacity. Opticalcomputing is a field with a bright future becauseof the natural characteristics of photons, such as parallelism, h
6、igh speed, and high bandwidth 1,2. Manyachievements of theory and engineering technologyhave been made on this subject during the past decades. The optical vectormatrix multiplier (OVMM)is a very important device in the optical computingdomain as a dedicated processor 3. Many studies4,5 focusing on
7、the algorithms based on vector andmatrix multiplication have been reported, which implies the OVMM can play important roles in severalfields.The OVMM is comprised of an optical system andelectrical circuits that provide the driving signal anddetect the output signal. The core processor of theOVMM is
8、 the optical system, which has been demonstrated in a variety of ways such as optical engineering 68, integrated optics 9, and nonlinear optics10,11. In this paper, an optical module of theOVMM based on optical engineering is proposed.It can transform the wavefront shape and control thepropagation p
9、ath of the light carrying informationwithout cross talk to implement a certain arithmeticoperation. In order to transform the beam shapeprecisely, the structure and the parameters of thesystem must be determined and optimized elaborately. They are also adjusted in consideration ofthe parameters of t
10、he existing commercial opticaldevices and electronic devices used in this system.Finally, the system should be fabricated accuratelyto ensure the imagery is formed at the correct location with acceptable aberrations.Basic Architecture and PrincipleOVMM is mainly used for matrix multiplication, vecto
11、r matrix multiplication, matrix multiplication operations. Its operation mode is Type of Stanford multiplier structure.34Vector matrix multiplication structure is functionally divided into: input vectors, a matrix light modulator, output vectors. The main role of the input unit of the light source i
12、s modulated to carry a uniform light beam to input to the information signal processing unit for each column of the array. The main role of the signal processing unit is input optical signals to achieve the multiplication of the control signals. The main role of the output unit is to converge the li
13、ght of each row of the array to the receiving unit and outputs the operation result signals.Traditional space optical vector matrix multiplier in the structure consists of the following parts:1) N Non-coherent light sources makes a source array.It is used for input vectors. Using a spherical lens an
14、d a cylindrical multiple lens to get optical spectral and spatial aggregation.2) The matrix that is to be multiplied is a light modulated matrix.3) After signal modulation matrix outputs through multiple sets of spherical lens and a cylindrical lens,put every line of light together to output vector
15、port that is composed of n light dectectors. Finally, we can output the calculation results though analog to digital converter(ADC).The above optical process model can be mathematically expressed as an m n matrix A and an n-dimensional vector obtained by multiplying a vector B to get a m-dimensional
16、 vector C. Take 2 2 matrix for example, input vector A(a1,a2) multiply the modulation matrix B and you will get the output vector C(c1,c2).As shown in table 2.1: There are some other theories and principles that are very useful and effective. Due to space limitations,they are not presented or you may see them in the following pages. Hotpots of theory and devi
copyright@ 2008-2022 冰豆网网站版权所有
经营许可证编号:鄂ICP备2022015515号-1