Showing papers by "Bell Labs published in 1978"

Variations of Box Plots

Abstract: Box plots display batches of data. Five values from a set of data are conventionally used; the extremes, the upper and lower hinges (quartiles), and the median. Such plots are becoming a widely used tool in exploratory data analysis and in preparing visual summaries for statisticians and nonstatisticians alike. Three variants of the basic display, devised by the authors, are described. The first visually incorporates a measure of group size; the second incorporates an indication of rough significance of differences between medians; the third combines the features of the first two. These techniques are displayed by examples.

Testing Software Design Modeled by Finite-State Machines

Abstract: We propose a method of testing the correctness of control structures that can be modeled by a finite-state machine. Test results derived from the design are evaluated against the specification. No "executable" prototype is required. The method is based on a result in automata theory and can be applied to software testing. Its error-detecting capability is compared with that of other approaches. Application experience is summarized.

Calculation of the Avrami parameters for heterogeneous solid state reactions using a modification of the Kissinger method

Abstract: Several isothermal experiments are generally needed to determine the parameters of the Avrami equation which describe most of the heterogeneous solid state reactions. Differential scanning calorimeters are suitable for such experiments. While most differential thermal analysis (DTA) apparatus cover a wider temperature range than DSC apparatus they cannot be used to perform isothermal determinations. However, Kissinger has already shown how activation energy and frequency factor can be calculated from DTA experiments for the case of homogeneous reactions following first order kinetics. We derive in this paper an extension of the Kissinger method and show its applicability to heterogeneous reactions described by an Avrami expression. The new method will allow the study of the kinetics of metallic reactions at the higher temperature range obtainable with DTA. The transformation kinetics of the metastable equiatomic tin-nickel alloy are given as an example.

Direct measurement of quasiparticle-lifetime broadening in a strong-coupled superconductor

Abstract: We have measured the quasiparticle recombination time in the strong-coupled superconductor ${\mathrm{Pb}}_{0.9}$${\mathrm{Bi}}_{0.1}$ directly by measuring the lifetime-broadened energy gap edge. This is done by measuring the $I\ensuremath{-}V$ characteristics of a superconducting tunnel junction of the type ${\mathrm{Pb}}_{0.9}$${\mathrm{Bi}}_{0.1}$-insulator-${\mathrm{Pb}}_{0.9}$${\mathrm{Bi}}_{0.1}$. Agreement with the calculated value is excellent.

Self-phase-modulation in silica optical fibers

Abstract: We report measurements of frequency broadening due to self-phase-modulation (SPM) in optical fibers. The use of single-mode silica-core fibers and mode-locked argon-laser pulses leads to the simplest and cleanest measurements yet made of SPM. The qualitative features of the frequency spectrum are in good agreement with theoretical expectations. The experiment provides an independent measurement of ${n}_{2}$, the self-focusing coefficient. The results also point to some simple and useful techniques based on fibers for the measurement and analysis of mode-locked laser pulses.

`` Strong '' NP-Completeness Results: Motivation, Examples, and Implications

Abstract: The NP-completeness of a computational problem ~s frequently taken to unply its \"mtractabthty\" However, there are certain NP-complete problems mvolvmg numbers, such as PARTITION and KNAPSACK, which are considered by many practitioners to be tractable The reason for this IS that, although no algontluns for solvmg them in tune bounded by a polynomial m the mput length are known, algorithms are known which solve them m tune bounded by a polynomial m the input length and the magmtude of the largest number an the given problem mstance. For other NP-complete problems mvolvmg numbers it can be shown that no such \"pseudopolynomml tune\" algonthra can exist unless P = NP. In this paper we provide a standard framework for stating and proving \"strong\" NP-completeness results of this sort, survey some of the strong NP-completeness results proved to date, and indicate some unphcauons of these results for both opumlzatlon and approximaUon algontluns

Asymptotic Theory of Least Absolute Error Regression

Abstract: In the general linear model with independent and identically distributed errors and distribution function F, the estimator which minimizes the sum of absolute residuals is demonstrated to be consistent and asymptotically Gaussian with covariance matrix ω2 Q -1, where Q = lim T -1 X'X and ω2 is the asymptotic variance of the ordinary sample median from samples with distribution F. Thus the least absolute error estimator has strictly smaller asymptotic confidence ellipsoids than the least squares estimator for linear models from any F for which the sample median is a more efficient estimator of location than the sample mean.

Dynamics of the charge-density wave. I. Impurity pinning in a single chain

Abstract: We study the effect of impurities on the dynamics of charge-density waves. At low temperature it is a good approximation to restrict our attention to phase fluctuation only. The competition between the random impurity potential and the elastic energy for phase fluctuations leads us naturally to consider two limiting cases. In the strong pinning case the phase is pinned to each impurity site. The pinning frequency and the frequency-dependent conductivity can be calculated. In the weak pinning case the pinning mechanism is more subtle. We show that the system is pinned by effectively breaking up into domains. The domain size is estimated which in turn determines the pinning frequency. The frequency-dependent conductivity and the static dielectric constant are calculated.

A charge-sheet model of the MOSFET

Abstract: Intuition, device evolution, and even efficient computation require simple MOSFET (metal-oxide-semiconductor field-effect transistor) models. Among these simple models are charge-sheet models which compress the inversion layer into a conducting plane of zero thickness. It is the purpose of this paper to test one such charge sheet model to see whether this approximation is too severe. This particular model includes diffusion which is expected to be important in the subthreshold and saturation regions. As a test the charge sheet model is applied to long-channel devices. Long-channel MOSFET behavior has been thoroughly studied, and is very well explained by the Pao-Sah double-integral formula for the current. Hence, a clear-cut test is a comparison of the charge sheet model with the Pao-Sah model. We find the charge sheet model has two advantages over the Pao-Sah model. (1) It leads to a very simple algebraic formula for the current of long-channel devices. The same formula applies in all regimes from subthreshold to saturation. Neither splicing nor parameter changes are needed. No discontinuities occur in either the current or the small-signal parameters, or in the derivatives of the small-signal parameters. (2) It is simpler to extend the charge sheet model to two or three dimensions than the Pao-Sah model. This simplification is a result of dropping the details of the inversion layer charge distribution. An important aspect of the gradual channel approximation is brought out by the analysis. Suppose the boundary condition relating the quasi-fermi level at the drain, φfL, to that at the source, φfo, namely φ ƒL =φ ƒ0 +V D where VD is the drain voltage, is applied in all bias regimes. Then it is shown that this means the potential at the drain end of the channel, φsL is not related to the potential at the source end of the channel, φso, by φ sL =φ s0 +V D Instead, φsL is computed, not imposed as a boundary condition. It is suggested that this failure of the potential to satisfy the boundary condition at the drain is justifiable. That is, φsL should be reinterpreted as the potential at the point in the channel where the gradual channel approximation fails. Hence, (2) may be relaxed. However, the “channel length” in the gradual-channel approximation now becomes a fitting parameter and is not the metallurgical source-to-drain separation. In addition several aspects of the long-channel MOSFET are brought out: (1) Pinch-off is achieved only asymptotically as the drain voltage tends to infinity. This is in marked contrast to the often-stated, textbook view that pinch-off is achieved for some finite drain voltage, the saturation voltage. (2) The channel or drain conductance approaches zero only asymptotically. (3) The transconductance saturates only asymptotically. Figures comparing the simple charge-sheet model formulas with the usual textbook formulas are included for direct-current vs drain voltage, channel conductance vs drain voltage, and transconductance vs drain voltage. The charge-sheet model agrees with the original Pao-Sah double-integral formula for the current at all gate and drain voltages, and possesses the correct subthreshold behavior. The textbook formulas do not.

Trapping of Atoms by Resonance Radiation Pressure

Abstract: A method of stably trapping, cooling, and manipulating atoms on a continuous-wave basis is proposed using resonance radiation pressure forces. Use of highly focused laser beams and atomic beam injection should give a very deep trap for confining single atoms or gases at temperatures \ensuremath{\sim} ${10}^{\ensuremath{-}6}$ \ifmmode^\circ\else\textdegree\fi{}K. An analysis of the saturation properties of radiation pressure forces is given.

Gaussian approximation of the fundamental modes of graded-index fibers

Abstract: Direct numerical integration of the wave equation is used to establish the validity of approximating the fundamental mode of graded-index fibers by a Gaussian function. We show that the fundamental modes of fibers, whose index profile can be expressed as a power law, are indeed very nearly Gaussian in shape (that is probably also true for graded-index fibers with convex profiles other than a power law). Graphs and empirical analytical expressions are presented for the optimum Gaussian beam width parameter and for the propagation constant of the fundamental mode.

The complexity of the network design problem

Abstract: In the network design problem we are given a weighted undirected graph. We wish to find a subgraph which connects all the original vertices and minimizes the sum of the shortest path weights between all vertex pairs, subject to a budget constraint on the sum of its edge weights. In this note we establish NP-completeness for the network design problem, even for the simple case where all edge weights are equal and the budget restricts the choice to spanning trees. This result justifies the development of enumerative optimization methods and of approximation algorithms, such as those described in a recent paper by R. Dionne and M. Florian.

The rate-distortion function for source coding with side information at the decoder\3-II: General sources

Abstract: In this paper we generalize (to nondiscrete sources) the results of a previous paper (Wyner and Ziv, 1976) on source coding with a fidelity criterion in a situation where the decoder (but not the encoder) has access to side information about the source We define R*(d) as the minimum rate (in the usual Shannon sense) required for encoding the source at a distortion level about d The main result is the characterization of R*(d) by an information theoretic minimization In a special case in which the source and the side information are jointly Gaussian, it is shown that R*(d) is equal to the rate which would be required if the encoder (as well as the decoder) is informed of the side information

A method for evaluating viscosities of metallic glasses from the rates of thermal transformations

Abstract: The apparent activation energies E(T) for the glass transition and crystallization in Pd77.6Cu6Si16.5 and Pd48Ni32P20 glass are seen to coincide with those for the viscous flow. This implies that both the rates of glass transition and crystallization in metallic glasses scale as the viscosity. Based on this proposition, the viscosity of a Pt45Ni30P25 glass, for example, has been evaluated, by means of thermal measurements, from the glass transition far into the crystallization temperatures. The viscosity η decreases rapidly from 1013 P at 480 K to 107 P at 580 K and is described by a Fulcher's expression as η = 10−3 exp[5950/(T − 320)].

A user-oriented software reliability model

Abstract: A user-oriented reliability model has been developed to measure the reliability of service that a system provides to a user community. It has been observed that in many systems, especially software systems, reliable service can be provided to a user when it is known that errors exist, provided that the service requested does not utilize the defective parts. The reliability of service, therefore, depends both on the reliability of the components and the probabilistic distribution of ulitization of the components to provide the service. In this paper a user-oriented reliability figure of merit is defined to measure the reliability of a software system with respect to a user environment. The effects of the user profile, which summarizes the characteristics of the users of a system, on system reliability is discussed. A simple Markov model is formulated to determine the reliability of a software system based on the reliability of each individual module and the measured inter-modular transition probabilities as the user profile. Sensitivity analysis techniques are developed to determine modules most critical to system reliability. The applications of this model to develop cost-effective testing strategies and to determine the expected penalty cost of failure are also discussed. Some future refinements and extensions of the model are presented.

Recombination enhanced defect reactions

Abstract: The energy liberated during an electronic transition may take the form of photons (luminescence), secondary electronic excitations (Auger), or phonons. In the case of phonon production, the vibrational excitation is localized at the recombination center (defect) before it dissipates to raise the temperature of the host lattice. A defect in an excited vibrational state may undergo simple solid state reactions such as diffusion, dissociation and annihilation which would not proceed under the Boltzmann distribution of energies determined by the temperature of the quiescent host lattice. The phenomenology of electronic enhancement is reviewed with emphasis on the roles of defect charge state, recombination and electric fields. The application of unimolecular reaction rate theory to the characterization of recombination enhanced reaction kinetics is discussed conceptually. Implications concerning the structural and chemical nature of host materials, recombination centers and reactant products is outlined. The role of recombination enhanced phenomena in device degradation is discussed and some useful applications of REDR in the selective programming of logic arrays and high definition electron beam data storage are presented.

Considerations in dynamic time warping algorithms for discrete word recognition

Abstract: The technique of dynamic time warping for time registration of a reference and test utterance has found widespread use in the areas of speaker verification and discrete word recognition. As originally proposed, the algorithm placed strong constraints on the possible set of dynamic paths-namely it was assumed that the initial and final frames of both the test and reference utterances were in exact time synchrony. Because of inherent practical difficulties with satisfying the assumptions under which the above constraints are valid, we have considered some modifications to the dynamic time warping algorithm. In particular, an algorithm in which an uncertainty exists in the registration both for initial and final frames was studied. Another modification constrains the dynamic path to follow (within a given range) the path which is locally optimum at each frame. This modification tends to work well when the location of the final frame of the test utterance is significantly in error due to breath noise, etc. To test the different time warping algorithms a set of ten isolated words spoken by 100 speakers was used. Probability density functions of the distances from each of the 100 versions of a word to a reference version of the word were estimated for each of three dynamic warping algorithms. From these data, it is shown that, based on a set of assumptions about the distributions of the distances, the warping algorithm that minimizes the overall probability of making a word error is the modified time warping algorithm with unconstrained endpoints. A discussion of this key result along with some ideas on where the other modifications would be most useful is included.

Data base communication call processing method

Abstract: A method is disclosed for processing call data among telephone switching offices and a remotely situated data base system for deriving call service information needed to complete special service calls, such as INward WATS (Wide Area Telephone Service) calls to a destination specified by the derived information. A network of local and toll offices compiles a special service call message including a caller dialed 800 number and an originating area code for the INWATS service and then transmits that message to the data base system over CCIS (Common Channel Interoffice Signalling) facilities including signal transfer systems. At the data base, the message is translated into a directory-unlisted number identifying an INWATS customer station designated to serve the call. Before the call is switched through the toll switching network under control of the unlisted number, the open/closed and busy/idle conditions of the INWATS station are checked at the data base to determine the availability of the station to serve the call. Traffic data is accumulated at the data base by originating area code and by the unlisted INWATS number. When the number of calls to the unlisted number per unit time exceeds a threshold, the data base instructs the telephone system to inhibit such calls from being served for a prescribed period.

Simple switching control method changes power converter into a current source

Abstract: A switching converter with an LC output filter behaves as a loose-tolerance voltage-controlled current source if each switch closure is ended when switch current reaches an adjustable threshold. This converter is then combined with an external feedback to produce a precise output voltage. By generating a fixed voltage with a current source in this manner, the converter has many advantages including continuous protection of the switches, stable and equal load sharing when several converters are operated in parallel, inherent overload protection, automatic switch symmetry correction, and fast system response.

Fast calculation of gain matrices for recursive estimation schemes

Abstract: A sequence of vectors {x(t)} with dimension N is given, such that x(t+1) is obtained from x(t) by introducing p new elements, deleting p old ones, and shifting the others in some fashion. The sequence of vectors $ is sought. This paper presents a method of calculating these vectors with proportional-to-Np operations and memory locations, in contrast to the conventional way which requires proportional-to-N 2 operations and memory locations. A non-symmetric case is also treated.

Core-Level Binding Energy and Density of States from the Surface Atoms of Gold

Abstract: X-ray photoemission spectra of $4f$ and valence electrons in surface atoms of gold have been obtained. The surface-atom $4f$ level is shifted 0.40\ifmmode\pm\else\textpm\fi{}0.01 eV to lower binding energy relative to the bulk value. The surface density of states is narrowed by (7.6 \ifmmode\pm\else\textpm\fi{} 1.1)% and its center of gravity is shifted to lower binding energy by 0.51\ifmmode\pm\else\textpm\fi{}0.08 eV. A model is proposed to account for the core-level shift in terms of the modified surface density of states.

Application of Fast Kalman Estimation to Adaptive Equalization

Abstract: Very rapid initial convergence of the equalizer tap coefficients is a requirement of many data communication systems which employ adaptive equalizers to minimize intersymbol interference. As shown in recent papers by Godard, and by Gitlin and Magee, a recursive least squares estimation algorithm, which is a special case of the Kalman estimation algorithm, is applicable to the estimation of the optimal (minimum MSE) set of tap coefficients. It was furthermore shown to yield much faster equalizer convergence than that achieved by the simple estimated gradient algorithm, especially for severely distorted channels. We show how certain "fast recursive estimation" techniques, originally introduced by Morf and Ljung, can be adapted to the equalizer adjustment problem, resulting in the same fast convergence as the conventional Kalman implementation, but with far fewer operations per iteration (proportional to the number of equalizer taps, rather than the square of the number of equalizer taps). These fast algorithms, applicable to both linear and decision feedback equalizers, exploit a certain shift-invariance property of successive equalizer contents. The rapid convergence properties of the "fast Kalman" adaptation algorithm are confirmed by simulation.

The Photorefractive Effect

Abstract: The optically induced change of the refractive index of electro-optic crystals, which was discovered over ten years ago in LiNb03, is now referred to as the photorefractive effect (by analogy with photochromism) Progress in our understanding of the microscopic mechanisms which has led to the development of optical recording sensitivities comparable to that of silver halide emulsions is reviewed Possibilities for application of the effect to optical memories, holographic interferometry, and integrated optics are considered

The attention operating characteristic: examples from visual search

Abstract: Even in the absence of eye movements, we show that subjects are able, upon instruction, to selectively attend to certain kinds of targets and parts of visual arrays. The major mechanism of altering attention is the switching of attention from trial to trial, although intermediate states of shared attention do occur. Attention operating characteristics are shown to be a useful way of describing such data and of assessing the compatibility of tasks to be performed simultaneously.

Counting large numbers of events in small registers

Abstract: It is possible to use a small counter to keep approximate counts of large numbers. The resulting expected error can be rather precisely controlled. An example is given in which 8-bit counters (bytes) are used to keep track of as many as 130,000 events with a relative error which is substantially independent of the number n of events. This relative error can be expected to be 24 percent or less 95 percent of the time (i.e. s = n/8). The techniques could be used to advantage in multichannel counting hardware or software used for the monitoring of experiments or processes.

Local moments and localized states

Abstract: What mathematical difficulty it contained has been almost fully- cleared up within the past few years. Localization was a different matter: very few believed it at the time, and even fewer saw its importance; among those who failed to fully understand it at first was certainly its author. It has yet to receive adequate mathematical treatment, and one has to resort to the indignity of numerical simulations to settle even the simplest questions about it. Only now, and through primarily Sir Nevill Mott’s efforts, is it beginning to gain general acceptance. Yet these two finally successful brainchildren have also much in common: first, they flew in the face of the overwhelming ascendancy. at the time of the band theory of solids, in emphasizing locality : how a magnetic moment, or an eigenstate, could be permanently pinned down in a given region. It is this fascination with the local and with the failures, not successes, of band theory, which the three of us here seem to have in common. Second, the two ideas were born in response to a clear experimental signal which contradicted the assumptions of the time; third, they intertwine my work with that of my two great colleagues with whom I have been jointly honored; and fourth, both subjects are still extremely active in 1977.

Visual discrimination of textures with identical third-order statistics

Abstract: We found a new class of two-dimensional random textures with identical third-order statistics that can be effortlessly discriminated. Discrimination is based on local "granularity" differences between these iso-trigon texture pairs. This is the more surprising since it is commonly assumed that texture granularity (grain) is determined by the power spectrum which, in turn, can be obtained from the second-order statistics. Because textures with identical third-order statistics must have identical second-order statistics (i.e., identical power spectra), visible texture granularity is not controlled by power spectra, and not even by third-order statistics.