Showing papers in "Bell System Technical Journal in 1979"

Advanced mobile phone service: The cellular concept

Abstract: This paper shows how a cellular system operating within a limited block of frequency spectrum can meet the objectives of a large-scale mobile-telephone service designed with attention to cost restraint It explores the key elements of the cellular concept — frequency reuse and cell splitting — and describes certain mathematical properties of hexagonal cellular geometry A description of the basic structure and features of amps shows how the cellular concept can be put into practice

Motion-compensated television coding: Part I

Abstract: We present methods of estimating displacements of moving objects from one frame to the next in a television scene and using such displacements for frame-to-frame prediction. Displacement is estimated by a recursive algorithm which seeks to minimize a functional of the prediction error. Several simplifications of the algorithm are presented which make it attractive for hardware implementation. Performance of the algorithm is evaluated by computer simulations on two sequences of moving images containing various amounts and types of motion. In both cases, the use of displacement-based (or motion-compensated) prediction results in bit rates that are 22 to 50 percent lower than those obtained by simple “frame-difference” prediction, which is used commonly in the interframe coders.

Multiqueue systems with nonexhaustive cyclic service

Abstract: Queuing models with cyclic-type service are applicable for performance studies of polling mechanisms in data communication and switching systems or cyclic scheduling algorithms in real time computers. This paper provides an approximate analysis of the multiqueue system M[x]/G/1 with batch Poisson input, general service times, general overhead (switchover) times, and a single server operating under a cyclic strategy with nonexhaustive service of queues. Based on a new concept of conditional cycle times, the generating function of the stationary probabilities of state, the Laplace-Stieltjes transforms of the delay distributions, and the mean waiting times are derived explicitly for each queue through an imbedded Markov chain approach and an independence assumption. The approximate analytic results are validated by computer simulations. Besides this analysis, a stability criterion is derived for the general case of GI/G/1 systems with cyclic priority service. The paper concludes with a number of studies of the behavior of cyclic queues discovering interesting properties such as the dependence of cycle times and waiting times on the arrival and service process types and on the efficiency of cyclic priorities.

A family of active switched capacitor biquad building blocks

Abstract: Two closely related, low-sensitivity, active switched capacitor fitter topologies are presented. Each of these circuits comprises two operational amplifiers and at most nine capacitors. The topologies have been carefully constructed so that they are immune to the various parasitic capacitances normally present in switched capacitor networks. One filter topology is capable of realizing any stable biquadratic z-domain transfer function, while the second one is only slightly less than fully general. Most commonly used transfer functions can be realized with either topology and will require only seven capacitors. The choice between the two topologies will generally be made on the basis of total capacitance required, dynamic range behavior, and sensitivity. A complete set of synthesis equations is given for both circuits which cover both the general and all the important special cases of the biquadratic transfer function. Finally, several examples are given which illustrate the synthesis procedures and the versatility of the filter topologies.

A new selective fading model: Application to propagation data

Abstract: Channel transmission models for use in estimating the performance of radio systems on line-of-sight paths at 6 GHz are explored. The basis for this study is the simple three-ray multipath fade, which provides a channel transfer function of the form H(ω) = a[1 − b exp −j(ω − ω 0 )τ], where a is the scale parameter, b is a shape parameter, τ is the delay difference in the channel, and ω 0 is the (radian) frequency of the fade minimum. This model is indistinguishable from an ideal channel model, within the accuracy of existing measurements. The propagation data that confirm the model were obtained in summer 1977 from a 26.4-mile hop near Atlanta, Georgia. The received power at 24 sample frequencies spaced at 1.1 MHz and centered on 6034.2 MHz was continuously monitored and recorded during periods of anomalous behavior. The model is applied to estimating the statistics of the channel delay difference, τ. The average delay difference giving rise to significant selectivity in the channel is between 5 and 9 ns. The distribution of delay difference is obtained for delay differences greater than 10 ns. The channel is found to have more than 3 dB of selectivity (difference between maximum and minimum attenuation in band) due to delay differences greater than 20 ns for more than 70 seconds in a heavy fading month. (This is comparable to the time the channel attenuation of a single frequency exceeds 40 dB.) The three-path model requires further simplification for narrowband channel application. For a channel with 30 MHz bandwidth, a model with fixed delay of 6.3 ns provides a sufficiently accurate representation of all observed channel conditions. The resulting nonphysical model is used to statistically characterize the condition of the fading channel. The statistics of the parameters of the fixed delay model are almost independent and of relatively simple form. The distribution of the shape parameter b is of the form (1 − b)2.3. The distribution of a is lognormal.

Determination of the basic device parameters of a GaAs MESFET

Abstract: This paper describes a new technique to determine the basic properties of the active channel of a gallium arsenide (GaAs) metal-semiconductor field effect transistor (mesfet). The effective gate length, channel thickness, and carrier concentration are determined from dc parameters. A precise method of measuring the dc parameters is also given. The new techniques are demonstrated using a wide variety of sample devices. It is also shown that microwave performance parameters, such as the maximum output power and minimum noise figure, are well predicted by dc parameters. Calculated values of the intrinsic and extrinsic dc parameters, using simple analytical expressions developed in terms of the geometrical and material parameters of a device, are shown to be in excellent agreement with their measured values. These expressions can be used as a basis for device design.

On the independence theory of equalizer convergence

Abstract: High-speed pulse amplitude modulated (pam) data transmission over telephone channels is only possible when adaptive equalization is used to mitigate the linear distortion found on the (initially unknown) channel. At the beginning of the equalization procedure, the tap weights are adjusted to minimize the inter symbol interference between pulses. The “stochastic gradient” algorithm is an iterative procedure commonly used for setting the coefficients in these and other adaptive filters, but a proper understanding of the convergence has never been obtained. It has been common analytical practice to invoke an assumption stating that a certain sequence of random vectors which direct the “hunting” of the equalizer are statistically independent. Everyone acknowledges this assumption to be far from true, just as everyone agrees that the final predictions made using it are in excellent agreement with experiments and simulations. We take the resolution of this question as our main problem When one begins to analyze the performance of the algorithm, one sees that the average mean-square error after the nth iteration requires knowing, as an intermediate step, the mathematical expectation of the product of a sequence of statistically dependent matrices. We transform the latter problem to a space of sufficiently high dimension where the required average may be obtained from a canonical equation V n+1 = A(α)V n + Here A(α) is a square matrix, depending on the “step-size” α of the original algorithm, and V n and F are vectors. The mean-square error is calculable from the solution V n .

Numerical integration of stochastic differential equations — ii

Abstract: In a previous paper, a method was presented to integrate numerically nonlinear stochastic differential equations (SDEs) with additive, Gaussian, white noise. The method, a generalization of the Range Kutta algorithm, extrapolates from one point to the next applying functional evaluations at stochastically determined points. This paper extends (and at one point corrects) algorithms for the simple class of equations considered in the previous paper. In addition, the method is expanded to treat vector SDEs, equations with time-dependent functions, and SDEs higher than first order. The parameters for several explicit integration schemes are displayed.

On the structure of real-time source coders

Abstract: The outputs of a discrete time source with memory are to be encoded (“quantized” or “compressed”) into a sequence of discrete variables. From this latter sequence, a receiver must attempt to approximate some features of the source sequence. Operation is in real time, and the distortion measure does not tolerate delays. Such a situation has been investigated over infinite time spans by B. McMillan. In the present work, only finite time spans are considered. The main result is the following. If the source is kth-order Markov, one may, without loss, assume that the encoder forms each output using only the last k source symbols and the present state of the receiver's memory. An example is constructed, which shows that the Markov property is essential. The case of delay is also considered.

Advanced mobile phone service: Introduction, background, and objectives

Abstract: This paper introduces a series of papers that describe in detail the Bell System's Advanced Mobile Phone Service (amps). It presents a brief history of mobile radio, highlighting the important events and legal decisions that preceded development of the amps system. The cellular system concept that has been embodied in amps makes large-scale mobile-radio service affordable to a sizable segment of the public. This concept calls for dividing transmission areas into “cells” to handle radio traffic, and, as traffic grows, subdividing those cells into smaller segments without increasing radio spectrum. This paper outlines amps objectives and sets the stage for more detailed articles on its evolution, its design and testing, and maintenance considerations.

On concentrators, superconcentrators, generalizers, and nonblocking networks

Abstract: In this paper, we study various communication networks, such as concentrators, superconcentrators, generalizers, and rearrangeable and nonblocking networks. We improve bounds for the number of edges (which can be viewed as approximations of the cost) in some networks by combinatorial analysis.

Imaging reflector arrangements to form a scanning beam using a small array

Abstract: To obtain the performance of a large aperture phased array, a small phased array is combined with a large main reflector and an imaging arrangement of smaller reflectors to form a large image of the small array over the main reflector. An electronically scanable antenna with a large aperture is thus obtained, using a small array. An attractive feature of the imaging arrangement is that the main reflector need not be fabricated accurately, since small imperfections can be corrected efficiently by the array. As an application, a 4.2-m diameter antenna is discussed for a 12–14 GHz satellite with a field of view of 3 degrees by 6 degrees required for coverage of the continental United States.

On the required tap-weight precision for digitally implemented, adaptive, mean-squared equalizers

Abstract: An analysis is made of the degree of precision required in a digitally implemented adaptive equalizer to achieve a satisfactory level of performance. Considering both the conventional synchronously spaced equalizer and the newer fractionally spaced equalizer, insight is provided into the relationship between the tap-weight precision and the steady-state, mean-squared error. It is demonstrated why the number of adaptive tap weights should be kept to a minimum (consistent with acceptable steady-state performance), both from convergence and precision requirements. A simple formula is given that displays the tradeoff among the equalizer mean-squared error, the number of taps, the channel characteristics, and digital resolution. For typical basic-conditioned voiceband channels operating at 9.6 kb/s, and neglecting the effects that limited resolution might have on timing and carrier phase tracking, analysis and simulation both indicate that the required top-weight resolution is of the order of 11 or 12 bits. Moreover, the minimum precision is only weakly dependent on the quality of the channel.

Effect of temperature on transmission in lightguides

Abstract: Temperature effects on refractive index have been studied by monitoring pulse propagation delays in fiber-optic lightguides for temperatures between −40 and 67°C. Loss and dispersion properties did not change significantly from their room temperature values.

Advanced mobile phone service: Control architecture

Abstract: The cellular concept used in the Advanced Mobile Phone Service (amps) system to achieve spectrum efficiency requires a complex and flexible distributed system control architecture. Three major subsystems serve as the control elements: the mobile unit, the cell site, and the switching office. System control functions are partitioned among these subsystems to handle the following amps control challenges: interfacing with the nationwide switched telephone network, dialing from mobile units, supervising calls from mobile subscribers in the presence of noise and co-channel interference, performing call setup functions including paging and access, and locating and handing off mobiles between cell sites. This paper explains the techniques used to achieve the control functions of the three major subsystems and the ways they in turn participate in control of the total amps system.

B.S.T.J. brief: Effects of sandstorms on microwave propagation

Abstract: Low rainfull volume suggests the promise of long paths using higher microwave frequencies for radio communication in desert areas. The pursuit of this promise gives rise to the need for understanding the effects of sandstorms on microwave propagation. First, a distinction should be made between large sand grains and fine sand dust1 Sand grains of greater than about O.2-mm diameter are driven by the wind as a low-flying cloud with a height of les than about 2 meters above the ground. This limited height is expected to be lower than most antenna heights of a microwave station. On the other hand, dust-like sand particles can rise in dense clouds to a height of one kilometer or more. This latter type of sandstorm, which is essentially a misnomer for dust storm, may lie in the terrestrial and earth-space paths of microwave radio; hence, path attenuation data are required. Precise calculation is hampered by the uncertainty about the dielectric constant and the size distribution of sand particles. However, useful analysis and frequency dependence of the sandstorm effects can be obtained without precise knowledge of these parameters.

Digital Radio outage due to selective fading — observation vs prediction from laboratory simulation

Abstract: A statistical model (introduced in a companion paper) of fading on a radio path is used with laboratory measurements on a digital radio system to estimate the outage due to multipath fading, where outage is the time that the bit error rate (ber) exceeds a threshold. Over the range of ber of interest (10−6 to 10−3), the calculated outage agrees favorably with the outage observed during the period for which the fading model was developed. It is further shown that the calculated outage, when scaled to a heavy fading month on the basis of single-frequency, time-faded statistics, agrees equally well with the outage observed on the same path during a heavy fading month The agreement between measured and predicted outage substantiates the selective fading model. The prescribed laboratory measurements characterize the sensitivity of the radio system to selective fading. Thus, the methodology provides a useful basis for comparing the outage of alternative realizations of digital radio systems.

Equivalent circuits for the analysis and synthesis of switched capacitor networks

Abstract: An extensive library of z-domain building block equivalent circuits is derived to facilitate the analysis and synthesis of switched capacitor (sc) networks. These sc building blocks, typically comprised of a single capacitor and from one to four switches, serve as basic circuit elements for sc networks in much the same spirit that resistors and capacitors serve analog networks. This building block library facilitates the derivation of canonic z-domain equivalent circuits for complex sc networks and the application of well-established mathematical network analysis and synthesis toots. What has been sought are easily applied techniques for achieving the same insights for sc networks that we have long enjoyed with analog networks.

B.S.T.J. brief: Optically powered speech communication over a fiber lightguide

Abstract: The photovoltaic conversion of optical power transmitted over a fiber lightguide can supply electrical power to low-drain semiconductor devices in remote locations. Acoustic powers comparable to those of conventional telephone ringers have been produced1 in this way by using a fiber-coupled GaALAs photovoltaic detector2 to excite an electroacoustic tone generator. It was conjectured1 that the electrical power for other telephone functions-transmit/receive, dialing, and hook-status recognition-could also be optically supplied, but the signaling techniques appropriate to a dielectric fiber were left unspecified. This note describes the implementation of two-way speech communication between an electrically powered local station and an optically powered station located at the remote end of a 1.1-km-long, single-strand, optical fiber. The remote-station sound alerter has also been operated over this link.

Current-access magnetic bubble circuits

Abstract: Experimental and theoretical results from our work on current-access technology show promise for high-density, ∼107 bits/cm2, and high-frequency, f 3 (FeGe) 5 O 12 , (LaLuSmCa) 3 (FeGe) 5 O 12 , and (LaLuSm) 3 (FeGa) 5 O 12 , the last appears best suited; some room temperature characteristics of the composition La 0.6 LU 2.1 Sm 0.3 Ga 0.9 Fe 4.1 O 12 are 4μMM, = 470G, μ = 750 cmS−1. Oe−1, d = 1.6 μm, and ΔH c = 2 Oe across a bubble for the threshold of motion. Necessary improvements in processing were made with a radio-frequency, chlorine-containing plasma etch which produced metal patterns identical to those of the etch mask. We anticipate that current-access devices, when compared to conventional field-access devices, will achieve higher data rates, lower power consumption per bit, and greater storage densities with existing processing technologies.

High resolution, steep profile, resist patterns

Abstract: High resolution and steep profile patterns have been generated in a 2.6-μm thick organic layer which conforms to the steps on a wafer surface and is planar on its top. This thick organic layer (a photoresist in the present experiments) is covered with an intermediate layer of SiO 2 and a top, thin layer of X-ray or photoresist. After exposure and development of the top resist layer, the intermediate layer is etched by CHF 3 reactive ion etching. The thick organic layer is then etched by O 2 reactive ion etching. Submicron resolution with essentially vertical walls in the thick organic material was achieved. The technique is also applicable to photo- and electron lithography. It reduces the need for thick resist patterns for the lithography step and, at the same time, ensures high resolution combined with good step coverage.

Variable rate coding of speech

Abstract: In this paper, we examine a number of concepts and issues concerning variable-rate coding of speech. We formulate the problem as a multistate coder (i.e., a coder that can operate at several bit rates) coupled with a time buffer. We first analyze the theoretical aspects of the problem by examining it in the context of a block processing formulation. We then suggest practical methods for implementing a variable rate coder based on a dynamic buffering approach. We also allude to a multiple user configuration of variable-rate coding for tasi-type applications. A practical example of a variable rate adpcm coder is presented and applied to speech coding. It is shown that by careful design the algorithm can be made to be as robust to channel errors as that of a fixed rate adpcm coder.

Motion-compensated transform coding

Abstract: Interframe hybrid transform/dpcm coders encode television signals by taking a spatial transform of a block of picture elements in a frame and predictively coding the resulting coefficients using the corresponding coefficients of the spatial block at the same location in the previous frame. These coders can be made more efficient for scenes containing objects in translational motion by first estimating the translational displacement of objects and then using coefficients of a spatially displaced block in the previous frame for prediction. This paper presents simulation results for such motion-compensated transform coders using two algorithms for estimating displacements. The first algorithm, which is developed in a companion paper, recursively estimates the displacements from the previously transmitted transform coefficients, thereby eliminating the need to transmit the displacement estimates. The second algorithm, due to Limb and Murphy, estimates displacements by taking ratios of accummulated frame difference and spatial difference signals in a block. In this scheme, the displacement estimates are transmitted to the receiver. Computer simulations on two typical real-life sequences of frames show that motion-compensated coefficient prediction results in coder bit rates that are 20 to 40 percent tower than conventional interframe transform coders using “frame difference of coefficients.” Comparisons of bit rates for approximately the same picture quality show that the two methods of displacement estimation are quite similar in performance with a slight preference for the scheme with recursive displacement estimation.

Combining echo cancellation and decision feedback equalization

Abstract: A system for two-wire, full-duplex data transmission is proposed. It consists of two adaptive transversal filters, one accepting the transmitted symbols and working as an echo canceller, the other accepting the received symbols and functioning as a decision feedback equalizer. A joint stochastic adjustment algorithm (updates at each baud) is analyzed, and it is shown that the sum of the mean-squared errors in the coefficients of both filters can be decoupled from its difference by selecting identical gain constants in each loop. The optimum gain equals the reciprocal of the sum of the taps of both loops. Convergence is exponential, and its time is 0.23 adjustments/dB/tap. This is completely independent of all channel parameters. Implementation of the proposed structure requires neither multipliers nor A/D converters. Promising applications are seen in channels with moderate precursor distortion, such as highpass channels (dcrestoration), two-wire pbx systems with a need for high-speed, full-duplex communication, limited distance cable channels, and, most important, two-wire digital subscriber lines for digital voice/data terminals.

Coefficient inaccuracy in transversal filtering

Abstract: Coefficient inaccuracy in transversal filters is known to degrade the frequency response, particularly in stopband regions. The magnitude of the stopband degradation increases with the number of stages n, the length of the impulse response. A widely used formula for the error in frequency response is proportional to √n. By extending recent results on random trigonometric polynomials, we show that for random additive coefficient errors with variance σ2, the error ΔH(ω) in frequency response for targe n is such that where log denotes the natural logarithm This result leads to an absolute bound on attainable stopband rejection for any band-select transversal filter with given coefficient inaccuracy. In particular, the result places a definite limitation on the quality of band-select filtering that can be achieved using a ccd split-electrode filter. It also implies bounds for the peak sidelobes of random radar arrays.

Automatic recognition of spoken spelled names for obtaining directory listings

Abstract: An automatic speaker-dependent word recognizer is used to accept strings of spelled letters spoken in isolation. The output of the recognizer is a set of best-candidate letters for each letter spoken in the string. Candidate strings forming spelled last names and initials are compared to name strings obtained from a telephone directory stored in a disk file. A systematic search is carried out to find matching entries in the directory. An evaluation has been carried out with ten talkers spelling lists of names extracted from an 18,000-entry telephone directory. Although the median acoustic error rate per spelled letter is approximately 20 percent, the median error rate in furnishing the requested directory entry is approximately 4 percent.

Transform domain motion estimation

Abstract: This paper introduces an algorithm for estimating the displacement of moving objects in a television scene from spatial transform coefficients of successive frames. The algorithm works recursively in such a way that the displacement estimates are updated from coefficient to coefficient. A promising application of this algorithm is in motion-compensated interframe hybrid transform- dpcm image coding. We give a statistical analysis of the transform domain displacement estimation algorithm and prove its convergence under certain realistic conditions. An analytical derivation is presented that gives sufficient conditions for the rate of convergence of the algorithm to be independent of the transform type. This result is supported by a number of simulation examples using Hadamard, Haar, and Slant transforms. We also describe an extension of the algorithm that adaptively updates displacement estimation according to the local features of the moving objects. Simulation results demonstrate that the adaptive displacement estimation algorithm has good convergence properties in estimating displacement even for very noisy images.

Three-stage multiconnection networks which are nonblocking in the wide sense

Abstract: A multiconnection network deals with the connections of pairs {(X, Y)} where X is a subset of the input terminals and Y is a subset of the output terminals. We study the conditions under which a three-stage Clos network is nonblocking for such connections. We show that the number of middle switches needed for nonblocking depends on the routing strategy. Therefore the networks satisfying the conditions are networks nonblocking in the wide sense. We also derive formulas for computing the minimum numbers of crosspoints required by such networks.

Rapid automatic index profiling of whole-fiber samples: Part II

Abstract: A new interferometric method not requiring any sample preparation is presented for determining the refractive-index profiles of optical fibers. The whole fiber is immersed in index-matching oil and placed transversely under an interference microscope. The determination of the refractive index distribution for an arbitrary, circularly symmetric fiber core requires the solution of an integral equation. In this first of two papers on the subject, we describe a method which accomplished its solution by assuming that the fiber core consists of a large number of concentric circular cylinders of step-wise constant refractive index. The index distribution can then be obtained by determining the index values of each layer successively. The method has been applied to double- and single-pass interferometric arrangements which are being compared here. The single-pass method is reproducible to about 1 percent and provides the complete index distribution within minutes with the help of a computer-controlled video-analysis system. The results of this method are in excellent agreement with profiles obtained from polished slabs of the same fiber.

Approximations to stochastic service systems, with an Application to a retrial model

Abstract: This paper illustrates the usefulness of state-dependent, birth-death processes in reducing the dimensions of stochastic service systems. The approximation techniques introduced have wide applicability to general (finite) multidimensional, state-dependent, birth-death processes. These techniques are introduced by considering the “classical” telephony problems dealing with trunk group overflow traffic from the point of view of state-dependent, birth-death processes. The main part of the paper then applies these techniques to a two-dimensional trunk group retrial model of Wilkinson and Radnik. The method, which reduces the W-R model to an approximate, easily-solved, one-dimensional model, makes use of the transition probabilities for state-dependent, birth-death processes. These are obtained via a simple extension of known results. We use the one-dimensional results to compute blocking for a range of parameter values (trunk group sizes and retrial rates) exceeding the computational limits of the W-R model. Maximum relative errors do not exceed 10 to 15 percent, while for most cases of practical interest the relative errors are less than 5 percent. The approximation also provides insight into the region of applicability of even simpler retrial models. This one-dimensional retrial model actually applies to more general (finite) state-dependent, birth-death processes (eg., loss-delay systems).