Parallel Networks for Machine Vision

The amount of computation required to solve many early vision problems is prodigious, and so it has long been thought that systems that operate in a reasonable amount of time will only become feasible when parallel systems become available. Such systems now exist in digital form, but most are large and expensive. These machines constitute an invaluable test-bed for the development of new algorithms, but they can probably not be scaled down rapidly in both physical size and cost, despite continued advances in semiconductor technology and machine architecture. Simple analog networks can perform interesting computations, as has been known for a long time. We have reached the point where it is feasible to experiment with implementation of these ideas in VLSI form, particularly if we focus on networks composed of locally interconnected passive elements, linear amplifiers, and simple nonlinear components. While there have been excursions into the development of ideas in this area since the very beginnings of work on machine vision, much work remains to be done. Progress will depend on careful attention to matching of the capabilities of simple networks to the needs of early vision. Note that this is not at all intended to be anything like a review of the field, but merely a collection of some ideas that seem to be interesting.

Analog neural nonderivative optimizers

Continuous-time neural networks for solving convex nonlinear unconstrained;programming problems without using gradient information of the objective function are proposed and analyzed. Thus, the proposed networks are nonderivative optimizers. First, networks for optimizing objective functions of one variable are discussed. Then, an existing one-dimensional optimizer is analyzed, and a new line search optimizer is proposed. It is shown that the proposed optimizer network is robust in the sense that it has disturbance rejection property. The network can be implemented easily in hardware using standard circuit elements. The one-dimensional net is used as a building block in multidimensional networks for optimizing objective functions of several variables. The multidimensional nets implement a continuous version of the coordinate descent method.

Analog nonderivative optimizers

Analog networks for solving convex nonlinear unconstrained programming problems without using gradient information of the objective function are proposed. The one-dimensional net can be used as a building block in multi-dimensional networks for optimizing objective functions of several variables.

An analog implementation of radial basis neural networks (RBNN) using BiCMOS technology

This paper describes a analog implementation of radial basis neural networks (RBNN) in BiCMOS technology. The RBNN uses a gaussian function obtained through the characteristic of the bipolar differential pair. The gaussian parameters (gain, center and width) is changed with programmable current source. Results obtained with PSPICE software is showed.

Bounds on Defect Level and Fault Coverage in Linear Analog Circuit Testing

Transfer function coefficients (TFC) are widely used to test linear analog circuits for parametric and catastrophic faults. This paper presents closed form expressions for an upper bound on the defect level (DL) and a lower bound on fault coverage (FC) achievable in TFC based test method. The computed bounds have been tested and validated on several benchmark circuits. Further, application of these bounds to scalable RC ladder networks reveal a number of interesting characteristics. The approach adopted here is general and can be extended to find bounds of DL and FC of other parametric test methods for linear and non-linear circuits.

Secure Transmission in Cooperative Networks with Weak Eavesdroppers

In this letter, we propose a scheme to improve the secrecy rate of cooperative networks using Analog Network Coding (ANC). ANC mixes the signals in the air; the desired signal is then separated out, from the mixed signals, at the legitimate receiver using techniques like self interference subtraction and signal nulling, thereby achieving better secrecy rates. Assuming global channel state information, memoryless adversaries and the decode-and-forward strategy, we seek to maximize the average secrecy rate between the source and the destination, subject to an overall power budget. Then, exploiting the structure of the optimization problem, we compute its optimal solution. Finally, we use numerical evaluations to compare our scheme with the conventional approaches.

On the Relationship Between the Generalized Equality Classifier and ART 2 Neural Networks

In this paper, we introduce the Generalized Equality Classifier (GEC) for use as an unsupervised clustering algorithm in categorizing analog data. GEC is based on a formal definition of inexact equality originally developed for voting in fault tolerant software applications. GEC is defined using a metric space framework. The only parameter in GEC is a scalar threshold which defines the approximate equality of two patterns. Here, we compare the characteristics of GEC to the ART2-A algorithm (Carpenter, Grossberg, and Rosen, 1991). In particular, we show that GEC with the Hamming distance performs the same optimization as ART2. Moreover, GEC has lower computational requirements than AR12 on serial machines.

Fuzzy ART: Fast Stable Learning and Categorization of Analog Patterns by an Adaptive Resonance System

A Fuzzy ART model capable of rapid stable learning of recognition categories in response to arbitrary sequences of analog or binary input patterns is described. Fuzzy ART incorporates computations from fuzzy set theory into the ART 1 neural network, which learns to categorize only binary input patterns. The generalization to learning both analog and binary input patterns is achieved by replacing appearances of the intersection operator (n) in AHT 1 by the MIN operator (Λ) of fuzzy set theory. The MIN operator reduces to the intersection operator in the binary case. Category proliferation is prevented by normalizing input vectors at a preprocessing stage. A normalization procedure called complement coding leads to a symmetric theory in which the MIN operator (Λ) and the MAX operator (v) of fuzzy set theory play complementary roles. Complement coding uses on-cells and off-cells to represent the input pattern, and preserves individual feature amplitudes while normalizing the total on-cell/off-cell vector. Learning is stable because all adaptive weights can only decrease in time. Decreasing weights correspond to increasing sizes of category "boxes". Smaller vigilance values lead to larger category boxes. Learning stops when the input space is covered by boxes. With fast learning and a finite input set of arbitrary size and composition, learning stabilizes after just one presentation of each input pattern. A fast-commit slow-recode option combines fast learning with a forgetting rule that buffers system memory against noise. Using this option, rare events can be rapidly learned, yet previously learned memories are not rapidly erased in response to statistically unreliable input fluctuations.

Fuzzy ARTMAP: A Neural Network Architecture for Incremental Supervised Learning of Analog Multidimensional Maps

A new neural network architecture is introduced for incremental supervised learning of recognition categories and multidimensional maps in response to arbitrary sequences of analog or binary input vectors. The architecture, called Fuzzy ARTMAP, achieves a synthesis of fuzzy logic and Adaptive Resonance Theory (ART) neural networks by exploiting a close formal similarity between the computations of fuzzy subsethood and ART category choice, resonance, and learning. Fuzzy ARTMAP also realizes a new Minimax Learning Rule that conjointly minimizes predictive error and maximizes code compression, or generalization. This is achieved by a match tracking process that increases the ART vigilance parameter by the minimum amount needed to correct a predictive error. As a result, the system automatically learns a minimal number of recognition categories, or "hidden units", to met accuracy criteria. Category proliferation is prevented by normalizing input vectors at a preprocessing stage. A normalization procedure called complement coding leads to a symmetric theory in which the MIN operator (Λ) and the MAX operator (v) of fuzzy logic play complementary roles. Complement coding uses on-cells and off-cells to represent the input pattern, and preserves individual feature amplitudes while normalizing the total on-cell/off-cell vector. Learning is stable because all adaptive weights can only decrease in time. Decreasing weights correspond to increasing sizes of category "boxes". Smaller vigilance values lead to larger category boxes. Improved prediction is achieved by training the system several times using different orderings of the input set. This voting strategy can also be used to assign probability estimates to competing predictions given small, noisy, or incomplete training sets. Four classes of simulations illustrate Fuzzy ARTMAP performance as compared to benchmark back propagation and genetic algorithm systems. These simulations include (i) finding points inside vs. outside a circle; (ii) learning to tell two spirals apart; (iii) incremental approximation of a piecewise continuous function; and (iv) a letter recognition database. The Fuzzy ARTMAP system is also compared to Salzberg's NGE system and to Simpson's FMMC system.

Secure Communication in Cellular Networks: The Benefits of Millimeter Wave Mobile Broadband

This paper proposes millimeter wave (mmWave) mobile broadband for achieving secure communication in downlink cellular network. Analog beamforming with phase shifters is adopted for the mmWave transmission. The secrecy throughput is analyzed based on two different transmission modes, namely delay-tolerant transmission and delay-limited transmission. The impact of large antenna arrays at the mmWave frequencies on the secrecy throughput is examined. Numerical results corroborate our analysis and show that mmWave systems can enable significant secrecy improvement. Moreover, it is indicated that with large antenna arrays, multi-gigabit per second secure link at the mmWave frequencies can be reached in the delay-tolerant transmission mode and the adverse effect of secrecy outage vanishes in the delay-limited transmission mode.

Monitoring and all-optical signal processing techniques for transparent networks

Optical networks are under constant evolution. The growing demand for dynamism require devices that can accommodate different types of traffic. Thus the study of transparent optical networks arises. This approach makes optical networks more "elegant" , due to a more efficient use of network resources. In this thesis, the author proposes devices that intend to form alternative approaches both in the state of art of these same technologies both in the fitting of this technologies in transparent optical networks. Given that full transparency is difficult to achieve with current technology (perhaps with more developed optical computing this is possible), the author proposes techniques with different levels of transparency. On the topic of performance of optical networks, the author proposes two techniques for monitoring chromatic dispersion with different levels of transparency. In Chapter 3 the proposed technique seems to make more sense for long-haul optical transmission links and high transmission rates, not only due to its moderate complexity but also to its potential moderate/high cost. However it is proposed to several modulation formats, particularly those that have a protruding clock component. In Chapter 4 the transparency level was not tested for various modulation formats, however some transparency is achieved by not adding any electrical device after the receiver (other than an analog-digital converter). This allows that this technique can operate at high transmission rates in excess of 100 Gbit / s, if electro-optical asynchronous sampling is used before the optical receiver. Thus a low cost and low bandwidth photo-detector can be used. In chapter 5 is demonstrated a technique for simultaneously monitoring multiple impairments of the optical network by generating novel performance analysis diagrams and by use of artificial neural networks. In chapter 6 the author demonstrates an all-optical technique for controlling the optical state of polarization and an example of how all-optical signal processing can fully cooperate with optical performance monitoring.

Supervised training approach using spiking neural networks

One of the basic aspects of some neural networks is their attempt to approximate as much as possible their biological counterparts. The goal is to achieve a simple and robust network, easy to understand and able of simulating the human brain at a computational level. Recently a third generation of neural networks (NN) [1], called Spiking Neural Networks(SNN) was appeared. This new kind of networks use the time of a electrical pulse, or spike, to encode the information. In the first and second generation of NN analog values are used in the communication between neurons.

Extensive Comparisons of Optical Fast-OFDM and Conventional Optical OFDM for Local and Access Networks

A performance comparison between a recently proposed novel technique known as fast orthogonal frequency-division multiplexing (FOFDM) and conventional orthogonal frequency-division multiplexing (OFDM) is undertaken over unamplified, intensity-modulated, and direct-detected directly modulated laser-based optical signals. Key transceiver parameters, such as the maximum achievable transmission capacity and the digital-to-analog/analog-to-digital converter (DAC/ADC) effects are explored thoroughly. It is shown that, similarly to conventional OFDM, the least complex and bandwidth efficient FOFDM can support up to similar to 20 Gb/s over 500 m worst-case multimode fiber (MMF) links having 3 dB effective bandwidths of similar to 200 MHz X km. For compensation of the DAC/ADC roll-off, a power-loading (PL) algorithm is adopted, leading to an FOFDM system improvement of similar to 4 dB. FOFDM and conventional OFDM give similar optimum DAC/ADC parameters over 500 m worst-case MMF, while over 50 km single-mode fiber a maximum deviation of only similar to 1 dB in clipping ratio is observed due to the imperfect chromatic dispersion compensation caused by one-tap equalizers.

A low-voltage low-power analog memory cell with built-in 4-quadrant multiplication

An accurate switched-current (SI) memory cell and suitable for low-voltage low-power (LVLP) applications is proposed. Information is memorized as the gate-voltage of the input transistor, in a tunable gain-boosting triode-transconductor. Additionally, four-quadrant multiplication between the input voltage to the transconductor regulation-amplifier (X-operand) and the stored voltage (Y-operand) is provided. A simplified 2 x 2-memory array was prototyped according to a standard 0.8 mum n-well CMOS process and 1.8-V supply. Measured current-reproduction error is less than 0.26% for 0.25 muA less than or equal to I-SAMPLE less than or equal to 0.75 muA. Standby consumption is 6.75 muW per cell @I-SAMPLE = 0.75 muA. At room temperature, leakage-rate is 1.56 nA/ms. Four-quadrant multiplier (4QM) full-scale operands are 2x(max) = 320 mV(pp) and 2y(max). = 448 mV(pp), yielding a maximum output swing of 0.9 muA(pp). 4QM worst-case nonlinearity is 7.9%.

An analog implementation of radial basis function network appropriate for transducer linearizer

A circuit for transducer linearizer tasks have been designed and built using discrete components and it implements by: a Radial Basis Function Network (RBFN) with three basis functions. The application in a linearized thermistor showed that the network has good approximation capabilities. The circuit advantages is the amplitude, width and center.