A monolithic 6 GHz quadrature frequency doubler with adjustable phase offset

A 6 GHz frequency doubler with quadrature outputs generated from a differential single phase input is presented. The phase offset between the in-phase and quadrature outputs can be digitally controlled in linear steps for use in an automated calibration algorithm. The design nominally achieves standard deviations in quadrature phase error and amplitude balance of 0.4/spl deg/ and 0.1dB, respectively. This is demonstrated with a single sideband (SSB) mixer that realizes an average uncalibrated sideband rejection of 48.2 dB which improves to 55.8 dB post-calibration under nominal conditions.     

Robust least-squares estimation with a relative entropy constraint

Given a nominal statistical model, we consider the minimax estimation problem consisting of finding the best least-squares estimator for the least favorable statistical model within a neighborhood of the nominal model. The neighborhood is formed by placing a bound on the Kullback-Leibler (KL) divergence between the actual and nominal models. For a Gaussian nominal model and a finite observations interval, or for a stationary Gaussian process over an infinite interval, the usual noncausal Wiener filter remains optimal. However, the worst case performance of the filter is affected by the size of the neighborhood representing the model uncertainty. On the other hand, standard causal least-squares estimators are not optimal, and a characterization is provided for the causal estimator and the corresponding least favorable model. The causal estimator takes the form of a risk-sensitive estimator with an appropriately selected risk sensitivity coefficient.     

Fault Isolation in Conventional Linear Systems?A Feasibility Study

A procedure for generating a set of tests for a conventional linear system making use of gain measurements from normal input to normal output terminals is described. Basically, the procedure is to compute the corner frequencies of the nominal network and choose, as test frequencies, several frequencies in the neighborhood of each corner frequency, and hence, detect shifts in the corner frequencies. Interpretations of measured values are obtained by precomputing the gain at these frequencies for parameter deviations from their nominal values. The theoretical basis of the procedure is also discussed. An example is worked out in detail to demonstrate the feasibility of the procedure.     

Stability in linear estimation

The stability with respect to model uncertainty of linear estimators of the coefficients of a linear combination of deterministic signals in noise is investigated. A class of estimators having nominal performances constrained to be close to that of the nominal linear, unbiased, minimum-variance (LUMV) estimator is specified. Two estimator stability indexes are defined, one based on a worst-case estimate mean-square error and the other on a type of signal-to-noise ratio. The estimator minimizing each index, subject to the optimality constraints, is found by reference to related LUMV estimation results. In most cases, the minimizing (or most stable) estimator is the same under the two indexes     

On weak convergence of probability measures, channel capacity andcode error probabilities

The relationship between weak convergence of channel probability measures, channel capacity, and error probability of block codes is examined for memoryless channels with general input and output alphabets. It is shown that channel capacity is a lower semi-continuous function and that every block code with maximal probability of error δ for a nominal channel for any ϵ>0 can be modified such that the modification has a probability of error less than δ+ϵ for all channels in a sufficiently small neighborhood of the nominal channel     

Fully implanted p-column InP field-effect transistor

P-column junction field-effect transistors (FET's) have been fabricated in semi-insulating InP utilizing an all-implanted planar technology. A periodic array of p+ columns implanted through an n-type channel layer is used to form the active region. Initial devices with nominal gate lengths of 4 µm exhibit a unity-power-gain frequencyf_{max}as high as 7 GHz while recent smaller gate length devices (nominally 1.7µm) show anf_{max}of 16 GHz.     

A fully scaled submicrometer NMOS technology using direct-write E-beam lithography

Fully scaled NMOS devices, circuits, and dynamic memory with 1/2-µm nominal minimum dimensions at each level have been fabricated using direct-write e-beam patterning. This high-density NMOS technology yields nominally loaded average gate delays of 650 ps/stage with a power dissipation of 38 µW. The characteristics of this technology are presented with specific emphasis placed on features of the design which are unique to submicrometer MOSFET's, including a study of nonscaling effects and their impact on the device and circuit design.     

Robust stability and sensitivity of input-output systems over extended spaces

This paper is the second part of [6] which is concerned with the sensitivity of general input-output systems over extended spaces. It is assumed that such systems, which need not be of feedback type, are governed by nonlinear operator equations relating the input, the state, and the output. These equations depend on a parameterA that can vary in a neighborhood of a nominal valueA ₀. Essentially, a system is called insensitive if any truncation of its output depends continuously onA provided the input is fixed. The theorems derived provide sufficient conditions for insensitivity. A control system of a feedback-feedforward type and a dynamical system described by a linear vector differential equation on [0, ) are discussed as examples.     

Nonlinear wavelet transforms for image coding via lifting

We investigate central issues such as invertibility, stability, synchronization, and frequency characteristics for nonlinear wavelet transforms built using the lifting framework. The nonlinearity comes from adaptively choosing between a class of linear predictors within the lifting framework. We also describe how earlier families of nonlinear filter banks can be extended through the use of prediction functions operating on a causal neighborhood of pixels. Preliminary compression results for model and real-world images demonstrate the promise of our techniques.     

Robust estimation and signal detection with dependent nonstationary data

In this paper the dual topics of robust signal detection and robust estimation of a random variable are considered, where the data may be both dependent and nonstationary. We note that classical saddlepoint techniques for robustness do not readily apply in the dependent and/or nonstationary situation, and thus our results have application in a larger domain than what was feasible heretofore. In addition, our methods make possible the quantitative measurement of robustness and admit essentially arbitrary perturbations in an underlying joint statistical distribution away from the nominal. In particular, our methods show that the presence of dependency can result in a reduction of the robustness of the linear detector by approximately 50% and that appropriate censoring can improve this situation. We also show that, somewhat surprisingly, a weak amount of censoring can actually reduce robustness rather than increase it, even with dependent data that is almost independent. This calls into question the common practice, inspired by classical saddlepoint results for independent data, of employing censoring in cases where residual dependency is conceded. When applied to estimation, our work shows that for nominally Gaussian data, the conditional expectation estimator is optimal not only in terms of performance but also robustness (under appropriate performance measures), thus reinforcing the appeal of this estimator. On the other hand, for other performance measures, we also note that the conditional expectation estimator can be completely unrobust, regardless of whether the data is nominally Gaussian or not. Finally, our results establish a bound on estimator robustness.This research was supported by the Air Force Office of Scientific Research under Grant AFOSR-91-0267.     

Robust Mixed$H_2/H_infty$Filtering for Time-Delay Fuzzy Systems

In this paper, a robust mixed$H_2/H_infty$filtering problem for continuous-time fuzzy systems subject to parameter uncertainties and multiple time-varying delays in state variables is addressed. The uncertain systems are expressed as Takagi–Sugeno fuzzy models with linear nominal parts and norm-bounded uncertainties. The main objective is to design stable filters that minimize a guaranteed cost index and achieve a prescribed$H_infty$performance under worst case disturbance. Based on Lyapunov theory, both delay-independent and delay-dependent sufficient conditions guaranteeing stability and achieving prescribed performances are stated in terms of linear matrix inequalities. Therefore, stable filters can be obtained easily with existing convex algorithms. Lastly, two examples are given to illustrate the proposed design methodology.     

Adaptive local linear regression with application to printer color management.

Local learning methods, such as local linear regression and nearest neighbor classifiers, base estimates on nearby training samples, neighbors. Usually, the number of neighbors used in estimation is fixed to be a global "optimal" value, chosen by cross validation. This paper proposes adapting the number of neighbors used for estimation to the local geometry of the data, without need for cross validation. The term enclosing neighborhood is introduced to describe a set of neighbors whose convex hull contains the test point when possible. It is proven that enclosing neighborhoods yield bounded estimation variance under some assumptions. Three such enclosing neighborhood definitions are presented: natural neighbors, natural neighbors inclusive, and enclosing k-NN. The effectiveness of these neighborhood definitions with local linear regression is tested for estimating lookup tables for color management. Significant improvements in error metrics are shown, indicating that enclosing neighborhoods may be a promising adaptive neighborhood definition for other local learning tasks as well, depending on the density of training samples.     

分段线性忆阻系统的簇发振荡及其机理分析

为了研究分段线性忆阻系统的簇发振荡及其形成机理,该文在一非自治系统中引入分段线性忆阻器模型与慢变化的周期激励项,建立了一种两时间尺度的4D分段线性忆阻系统。由于分段线性忆阻器模型的引入,系统被非光滑分界面分成不同的子系统。相应子系统控制的名义平衡轨迹的稳定性与非光滑分界面均会影响系统的簇发现象,导致轨迹在非光滑分界面处的突然跃迁与非光滑分岔的产生,从而展现出两种不同机理的簇发模式。利用微分包含定理对分岔机理进行分析,并借助时序图、转换相图等,通过数值仿真与Multisim电路仿真验证了理论分析的正确性,该文对分段线性忆阻系统的动力学行为及应用研究具有重要意义。     

Performance Evaluation of a Virtual Queuing – Channel Allocation Scheme for Microcellular Networks

This paper deals with the performance evaluation of a novel distributed channel allocation scheme that attains a high resource reuse in cellular networks. A suitable interference model has been envisaged. Resources are nominally assigned to cells with Fixed Channel Allocation (FCA), according to a given reuse distance. Whenever a channel demand from a mobile user does not find a free nominal resource in a cell, a channel is temporarily borrowed violating the reuse distance constraint, provided that the co-channel interference level is acceptable. As soon as a nominal channel becomes available in this cell, the borrowed resource is released (virtual queuing). The performance of the proposed channel allocation scheme has been evaluated considering both uniform and non-uniform traffic patterns. A performance analysis approach has been also developed and validated by simulations. Comparisons with the classical FCA scheme and other dynamic and borrowing channel allocation strategies have permitted to highlight the good performance of the proposed technique.     

When is the stability of a nonlinear input-output system robust?

We consider a general nonlinear input-output system governed by operator equations that relate the system's input, state, and output, all of which are in extended spaces. It is assumed that the system variables are separated. Our results give conditions under which the stability of the nominal system is robust; i.e., it is not destroyed by any sufficiently small admissible perturbation of the system. Theorem 1 deals with the case when by stability we mean theincremental stability. Theorem 3 concerns the^*-stability; i.e., the case when the stability is essentially the boundedness of the transmission operator. Moreover, in Theorem 2 it is shown that, under certain conditions, the incremental stability of the nominal system implies insensitivity. Basically, our results show that if the operators describing the nominal system are well behaved, and the transition from the nominal system to the perturbed system is not abrupt, then the nominal system stability is robust. The applications of the results are illustrated by several examples.     

Robust stability and senstivity of input-output systems over extended spaces

In Part I of this paper we consider a general model of an input-output system governed by nonlinear operator equations that relate the system's input, state, and output, all of which are elements in extended spaces. This model encompasses feedback systems as a special case. Assuming that the equations governing the system depend on a parameterA that is allowed to vary in a neighborhoodN _r (A ₀₎ of a nominal valueA ₀ in a linear space, we study conditions under which the system is stable for everyA ε N_r(A₀), i.e., when the system exhibits robust stability. By stability we essentially mean that the input-output operator is continuous. Depending on the type of continuity of a map between two extended spaces, four concepts of robustness are introduced. The main results, Theorems 1 and 2, furnish sufficient conditions for a system to be robust in the respective sense. Basically, they show that if the nominal system satisfies a certain condition guaranteeing its stability, and the operators appearing in the governing equations depend continuously on the parameterA, then we have robust stability. As examples illustrating the applications of our results we discuss (1) a feedback-feedforward system, in particular the case when the extended space consists of locally square-integrable functions or functions continuous on [0, ∞), and (2) a time-varying dynamical system described by a linear vector differential equation, whose variables are continuous functions on [0, ∞) which decrease exponentially to zero ast → ∞. At the end of the paper some modifications of the presented theory are discussed.     

Adaptive incremental sliding mode control for a robot manipulator

An adaptive incremental sliding mode control (AISMC) scheme for a robot manipulator is presented in this paper. Firstly, an incremental backstepping (IBS) controller is designed using time-delay estimation (TDE) to reduce dependence on the mathematical model. After substituting IBS controller into the nonlinear system, a linear system w.r.t. tracking errors is obtained while TDE error is the disturbance. Then, the AISMC scheme, including a nominal controller and an SMC, is developed for the resulted linear system to improve control performance. According to the equivalent control method, the SMC in the AISMC scheme is to handle TDE error. To receive optimal control performance at the sliding manifold, an LQR controller is selected as the nominal controller. The SMC is designed based on positive semi-definite barrier function (PSDBF) since it prevents switching gains from being over/under-estimated, and two practical problems are addressed in this paper: A new PSDBF is designed and conservative (large) setting bounds affecting tracking precision and/or system stability are avoided; An improved PSDBF based SMC is developed where the PSDBF and an adaptive parameter are used simultaneously to regulate switching gains, and the system is still stable when sliding variable occasionally exceeds the predefined vicinity. Moreover, finite-time convergence property of the sliding variable is strictly analyzed. Finally, real-time experiments are conducted to verify the effectiveness of the proposed control method.     

Critical stability constraints for discrete-time linear systems

Critical stability constraints are a small set of conditions that are enough to maintain the stability of a system when some parameters are perturbed from a nominal stable setting. The paper uses a recently introduced efficient integer-preserving (IP) form of the Bistritz test to derive critical constraints for stability of discrete-time linear systems. The new procedure produces polynomial (rather than rational) constraints of particularly low degree whose variates are the free parameters (or the literal coefficients) of the system's characteristic polynomial. Comparison with the modified Jury test, also an efficient IP stability test, shows that the constraints are obtained with less computation and, more contributing to simplicity, the constraints appear as polynomials with degrees lower by a factor of two.     

Comparison of Nd 1.06 and 1.33 µm operation in various hosts

Performances of Nd in various hosts including Nd:YAG, Nd:YLF, Nd:BEL, and Nd:Cr:GSGG were characterized and compared for operation on the nominal 1.06 μm and 1.33 μm transitions. Comparison included continuous mode operation, normal mode operation,Q-switched operation, and gain measurements. Continuous lasing of Nd:Cr:GSGG under lamp pumping is reported for what may be the first time. Characterization included operation with different mirror reflectivities to obtain the internal losses, measurement of the upper laser level lifetime, measurement of the beam quality of the lasers operating under nominally similar circumstances, and measurement of the wavelength. These measurements should provide a comparison of the various Nd hosts under similar circumstances for a wide variety of operational modes and at both wavelengths of interest.     

Low-Complexity Estimation of the Nominal Azimuth and Elevation for Incoherently Distributed Sources

This paper develops a new technique for estimating the two-dimensional direction-of-arrivals (DOAs) of incoherently distributed (ID) sources, which can estimate effectively the nominal azimuth and nominal elevation of multiple ID sources at the cost of less computational complexity. Using a pair of parallel uniform linear arrays (ULAs), a new approach for 2D DOA estimation of multiple ID sources is proposed. The proposed method firstly estimates the nominal elevation by the modified TLS-ESPRIT method, which is based on the approximate rotational invariance property with respect to the nominal elevation between two closely parallel ULAs. And then with the help of the nominal elevation estimates, the nominal azimuth is estimated by one-dimensional searching. Without multi-dimensional searching, the proposed method has significantly reduced the computational cost compared with the existing methods. Simulation results indicate that the proposed method can exhibit a good performance and be applied to the multisource scenario where different sources may have different angular distribution shapes.