| By the end of 1960s, since the conventional optics system having the big volume, poor stability and difficulty of beam collimation could not satisfy the requirements of optical communication and optical information processing, people hoped to realize integrated optics circuits like the molectron. With the effort on it for several decades, some research productions have already displayed the important functions in the fields of communication, military, power, astronomy, sensors, etc. At the same time, integrated optics has been formed as a new subject with the cross-link of optics and pellicular electronics. The applied fields of integrated optics exist in many aspects. Besides optical communication, sensors, optical information processing and optical computers, they are still infiltrating into other fields, such as material science, optical instrument, spectrum research, etc. According to the emonstration effection of the integrated electronics, many scientists chose the potential researches and developed many kinds of integrated optical devices. Integrated optics, like the integrated electronics, will make a profound change of the information technology. Optical waveguides are the base of the integrated optical circuits. They are the carriers for rapidly transmitting optical signals in large quantities. As the elements for restricting and guiding the lights, their essential is that the refractive index of waveguides is higher than that of external medium. In the integrated optical devices, structure characteristics and transmission characteristic of channel waveguides are similar with ones of single-mode fiber. Compared with other waveguides, channel waveguides coupled more easily with single-mode fiber and have less coupler loss. However, because of the complex fabrication techniques of channel waveguides circuits, the application of them is still within the laboratory limits. In order to put the channel waveguides device into practical use, the measurement of parameters and characters becomes very important, especially the measurement of the refractive index profile (RIP). Considering this point, we devoted to find a good method to measure the RIP of optical waveguides. In the thesis, the involved contents and present state of waveguides are firstly introduced. They include the structure characteristics, the index profile features, preparing methods, mode eigenvalue equation and approximate treatment methods. In addition, it is summarized that some popular technology about the measurement of RIP. The main research works concentrate on the construction to the RIP of optical waveguides from Near-Field (NF) theology with an inverse algorithm method. Here, single-mode fiber as the measured object which RIP has been constructed, in order to not only prove feasibility of this technology, but also supplynecessary references for preparing the channel waveguides'. For three-dimensional optical waveguides, the refractive-index differenceΔn between guided wave area and substrate medium is determined by (,)(,)2?n =n( x,y)?ns ≈?1n? s kt2 φ2φxxyy (1) where φ( x ,y)=I(x,y)Imax is the transverse electric or magnetic field, I(x,y) is the measured intensity profile and Imax is the maximum of I(x,y), n(x,y) is the refractive-index profile, ns = βk is the cladding index of fiber, βis the propagation constant, k is the wave number in vacuum. From Eq.(1), we are aware that the RIP of the measured optical waveguide is determined by optical field information on its end surface. Based on this thought, we designed a suit of the experiment setup for the RIP measurement, as shown in Fig.1. In this graph, the measured object is single-mode fiber.However, the NF method requires that the measurement results of optical intensity profile have a high precision, so that it is necessary to filter the obtained pattern. The traditional filtering methods need multi-iteration and smoothing, which not only increases the number of calculation and reduces the efficiency of measurement, but also affects on the accuracy of results. In this thesis, we used an inverse algorithm mathematical treatment method to accomplish filtering through the simple matrix transform and the problem was solved in linear domain. With the simplification of one dimension, according to the theory of the finite-difference method, Eq.(1) can be expressed by sjjj jjnknφ?1 ? ?2φy 2 +φ+1=?22?φ(2a) the boundary conditions φ0 = φ(0), φn +1 =φ(n?y) (2b) where φi = φ( j?y). In this equation, the subscripts j and n are the discrete grid point and the number of computation points along y axis, ?y is the data points spacing along y axis, n?y is the core diameter of the fiber. Taking matrix transformation and the least-square method for Eq.(2),we can obtain the simple linear relation between the immediate measured quantity φmeasured and the unknown quantity N estimatedwhich has been corrected: measuredmeasuredTTN estimated = (? ?)?1?φ=Rφ(3)With the simplification of one dimension, the RIP of optical waveguides can be constructed from the NF pattern on the end surface, combined with an inverse algorithm filtering method, as shown in Fig.2. If we consider that 2-D optical field affects on the RIP, two results of the RIP calculated from optical intensity measured from horizontal and vertical directions need to be averaged. To demonstrate and evaluate the technique, we compared two these results that measured the RIP in the condition of simplification of one dimension and influences of the 2-D optical field, respectively, as shown in Fig.3. The result demonstrates that the RIP calculated from 2-D field information is more exact. The method has advantages on convenient operation and accurate results for the characterization of optical waveguides. It can be applied to the RIP measurement of single-mode fiber, channel waveguides and single-mode...
|
PDF Full Text Request