The dependence of the negative nonlinear absorption effect on the modulation degree and frequency of the incident laser was investigated in Er3+: LiYF4 crystals. With a decreasing modulation degree, a reverse-phased waveform was obtained in the transmitted waveform for modulation intensities of sinusoidal and rectangle waves. The transmitted waveform was observed symmetrically at modulation frequency of 0. 25 MHz. However, the transmitted waveform was asymmetrical at higher than 0. 5 MHz. In addition, the reverse-phased waveforms were obtained for sample lengths greater than 12 mm in the incident modulation degree of 10 to 56%. The NNA effect was observed at a temperature range of 20 to 500 K. It has been confirmed that a mechanism of the negative nonlinear absorption can be explained by considering an enhanced absorption model for a five-level system of the Er3+ ion.

We demonstrate a very simple and compact optical transceiver diode module using a passive alignment on a silicon bench with a V-groove. The excess loss caused by the passive alignment of an optical transceiver diode and a flat-end optical fiber is only 0. 6 dB. A high coupling efficiency of -4. 3 dB is obtained. This results in a high responsivity with a wavelength- and polarization-independence of 0. 5 dB over a 70 nm wavelength range and in good laser performance.

This paper proposes an object oriented face region detection and tracking method using range color information. Range segmentation of the objects are obtained from the complicated background using disparity histogram (DH). The facial regions among the range segmented objects are detected using skin-color transform technique that provides a facial region enhanced gray-level image. Computationally efficient matching pixel count (MPC) disparity measure is introduced to enhance the matching accuracy by removing the effect of the unexpected noise in the boundary region. Redundancy operations inherent in the area-based matching operation are removed to enhance the processing speed. For the skin-color transformation, the generalized facial color distribution (GFCD) is modeled by 2D Gaussian function in a normalized color space. Disparity difference histogram (DDH) concept from two consecutive frames is introduced to estimate the range information effectively. Detailed geometrical analysis provides exact variation of range information of moving object. The experimental results show that the proposed algorithm works well in various environments, at a rate of 1 frame per second with 512 480 resolution in general purpose workstation.

We propose an information integration architecture for a man-machine interface to construct a new agent-based Computer Supported Cooperative Work (CSCW) system. The system acts as a clerk in cooperative work giving users the advantage of using cooperative work space. The system allows users to do their work in the style of an ordinary meeting because spontaneous expressions of speech and gestures by users are detected by sensors so that they can be integrated with a task model at several levels to create suitable responses in a man-machine interface. As a result, users can dedicate themselves to mutually understand other meeting members with no awareness of direction to the CSCW system. In this paper, we describe the whole system and its information integration architecture for the man-machine interface including, the principle of functions, the current status of the system and future directions.

This paper proposes a simple learning algorithm that can realize any boolean function using the three-layer binary neural networks. The algorithm has flexible learning functions. 1) moving "core" for the inputs separations,2) "don't care" settings of the separated inputs. The "don't care" inputs do not affect the successive separations. Performing numerical simulations on some typical examples, we have verified that our algorithm can give less number of hidden layer neurons than those by conventional ones.

This paper describes a real-time connection admission control scheme for supporting multiple service categories. The scheme is based on a real-time cell-loss ratio evaluation algorithm for VBR based on peak/sustainable cell rates and maximum burst size. The algorithm is based on a notion of Allan variance of VP utilization. The most remarkable characteristics of the admission control scheme are that it terminates within constant time, a few milliseconds, and that its time is independent of both the number of VCs and the capacity of a cell buffer.

In the IETF, discussions on the authentication method of the Dynamic Host Configuration Protocol (DHCP) message are active and several methods have been proposed. These related specifications were published and circulated as the IETF Internet-Drafts. However, they still have several drawbacks. One of the major drawbacks is that any user can reuse addresses illegally. A user can use an expired address that was allocated to a host. This kind of "illegal use" of the addresses managed by the DHCP server may cause serious security problems. In order to solve them, we propose a new access control method to be used as the DHCP message authentication mechanism. Furthermore, we have designed and developed the DAG (DHCP Access Control Gateway) according to our method. The DAG serves as a gateway that allows only network accesses from clients with the address legally allocated by the DHCP server. This provides secure DHCP service if DHCP servers do not have an authentication mechanism, which is most likely to occur. If a DHCP server has such an authentication scheme as being proposed in IETF Internet-Draft, the DAG can offer a way to enable only a specific client to access the network.

An asynchronous pulse neural network model which is suitable for VLSI implementation is proposed. The model neuron can function as a coincidence detector as well as an integrator depending on its internal time-constant relative to the external one, and show complex dynamical behavior including chaotic responses. A network with the proposed neurons can process spatio-temporal coded information through dynamical cell assemblies with functional synaptic connections.

This paper presents both new analytical and new numerical solutions to the problem of generating waveforms exhibiting a low peak-to-peak factor. One important application of these results is in the generation of pseudo-white noise signals that are commonly uses in multi-frequency measurements. These measurements often require maximum signal-to-noise ratio while maintaining the lowest peak-to-peak excursion. The new synthesis scheme introduced in this paper uses the Discrete Fourier Transform (DFT) to generate pseudo-white noise sequence that theoretically has a minimized peak-to-peak factor, Fp-p. Unlike theoretical works in the literature, the method presented here is based in purely discrete mathematics, and hence is directly applicable to the digital synthesis of signals. With this method the shape of the signal can be controlled with about N parameters given N harmonic components. A different permutation of the same set of offset phases of the "source harmonics" creates an entirely different sequence.

In order to improve microprocessor performance, we propose to utilize histories of dynamic instruction sequences. A lot of special purpose memories integrated in a processor chip hold the histories. In this paper, we describe the usefulness of using two special purpose memories: Non-Consecutive basic block Buffer (NCB) and Reference Prediction Table (RPT). The NCB improves instruction fetching efficiency in order to relieve control dependences. The RPT predicts data addresses in order to speculate data dependences. From the simulation study, it has been found that the proposed mechanisms improve processor performance by up to 49. 2%.

Though response of neurons is mainly decided by synaptic events, the length of a time window for the neuronal response has still not been clarified. In this paper, we analyse the time window within which a neuron processes synaptic events, on the basis of the Hodgkin-Huxley equations. Our simulation shows that an active membrane property makes neurons' behavior complex, and that a few milliseconds is plausible as the time window. A neuron seems to detect coincidence synaptic events in such a time window.

Conventional approaches to neural network training do not consider possibility of selecting training samples dynamically during the learning phase. Neural network is simply presented with the complete training set at each iteration of the learning. The learning can then become very costly for large data sets. Huge redundancy of data samples may lead to the ill-conditioned training problem. Ill-conditioning during the training causes rank-deficiencies of error and Jacobean matrices, which results in slower convergence speed, or in the worst case, the failure of the algorithm to progress. Rank-deficiencies of essential matrices can be avoided by an appropriate selection of training exemplars at each iteration of training. This article presents underlying theoretical grounds for dynamic sample selection (DSS), that is mechanism enabling to select a subset of training set at each iteration. Theoretical material is first presented for general objective functions, and then for the objective functions satisfying the Lipschitz continuity condition. Furthermore, implementation specifics of DSS to first order line search techniques are theoretically described.

Simplification orderings, like the recursive path ordering and the improved recursive decomposition ordering, are widely used for proving the termination property of term rewriting systems. The improved recursive decomposition ordering is known as the most powerful simplification ordering. Recently Jouannaud and Rubio extended the recursive path ordering to higher-order rewrite systems by introducing an ordering on type structure. In this paper we extend the improved recursive decomposition ordering for proving termination of higher-order rewrite systems. The key idea of our ordering is a new concept of pseudo-terminal occurrences.

This paper describes a new architecture-based DSP processor, which consists of n n mesh multiprocessor for digital signal processing. A prototype chip, RCDSP9701 has been designed and implemented using a CMOS 0. 6 µm process. This architecture has better performance compare to the traditional microprocessor solution to Digital Signal Processing. The proposed method poses remarkable flexibility compare to ASIC (Application Specified Integrated Circuits) approach for Digital Signal Processing applications. In addition, the proposed architecture is fault tolerant and suitable for parallel computing applications. In this paper, an implementation into a silicon chip of the new architecture is presented to give a better understanding of our work.

A reconfiguration method for processor array is proposed in this paper. In the method, genetic algorithm (GA) is used for searching effective spare arrangement, which leads to successful reconfiguration. The effectiveness of the method is demonstrated by computer simulations.

Computational expensiveness of the training techniques, due to the extensiveness of the data set, is among the most important factors in machine learning and neural networks. Oversized data set may cause rank-deficiencies of Jacobean matrix which plays essential role in training techniques. Then the training becomes not only computationally expensive but also ineffective. In [1] the authors introduced the theoretical grounds for dynamic sample selection having a potential of eliminating rank-deficiencies. This study addresses the implementation issues of the dynamic sample selection based on the theoretical material presented in [1]. The authors propose a sample selection algorithm implementable into an arbitrary optimization technique. An ability of the algorithm to select a proper set of samples at each iteration of the training has been observed to be very beneficial as indicated by several experiments. Recently proposed approaches to sample selection work reasonably well if pattern-weight ratio is close to 1. Small improvements can be detected also at the values of the pattern-weight ratio equal to 2 or 3. The dynamic sample selection approach, presented in this article, can increase the convergence speed of first order optimization techniques, used for training MLP networks, even at the value of the pattern-weight ratio (E-FP) as high as 15 and possibly even more.

Two families of rapidly synchronizable spreading codes are compared using the same component codes. The influence of component code choice is also discussed. It is concluded that correlation, code-division multiple-access (CDMA) and information security (measured by the value of linear complexity) properties of Kronecker sequences are considerably better than those of Combination sequences. Combination sequences cannot be recommended for CDMA use unless the number of active users is few. CDMA performance of Kronecker sequences is almost comparable with that of linear pseudonoise (PN) code families of equal length when a Gold or Kasami code is used as the innermost code and the Barker code is used as the outermost code to guarantee satisfactory correlation and CDMA properties. Kronecker sequences possess a considerably higher value of linear complexity than those of the corresponding non-linear Geffe and majority logic type combination sequences. This implies they are highly non-linear codes due to the Kronecker product construction method. It is also observed that the Geffe type Boolean combiner resulted in better correlation and CDMA performance than with majority logic. The use of the purely linear exclusive-or combiner for considerable reduction of code synchronization time is not found recommendable although it results in good CDMA performance.

The proposed method tracks the user location in a hierarchically distributed fashion. Call patterns in personal communications services (PCS) have the locality. The virtually hierarchical link (VHL) is used to support the call locality and makes the fast call set up possible. The locality consideration effectively reduces the query traffic compared with non-locality consideration.

As a data transmission rate must be increased as required to support the future high-speed wireless communication systems under multipath fading, the conventional DS-CDMA scheme suffers considerably from an intensive processing requirement for the increased spreading rate to combat the inter-chip interference (ICI) and furthermore, from the intersymbol interference (ISI) as the symbol duration becomes less than the channel delay spread. In this paper, a multi-carrier parallel combinatory DS-CDMA (MC-PC-CDMA) scheme is considered as one possible variant access scheme to realize a bandwidth efficient transmission for high transmission rate while maintaining the beneficial features of the DS-CDMA scheme. This scheme combines the parallel combinatory signaling feature of the existing parallel combinatory CDMA (PC-CDMA) scheme with the orthogonal carrier multiplexing feature of multi-carrier modulation so as to improve the bandwidth efficiency and to reduce the self-interference among the parallel spreading sequences of each user, respectively. This particular system configuration also treats the previously proposed multi-carrier DS-CDMA systems as a special case. Our analysis of the bit error rate for the asynchronous CDMA system investigates the performance characteristics of the proposed system on varying design parameters, and shows the performance comparison with other types of multi-carrier DS-CDMA systems.

In the design of nonlinear reliable controllers, one major issue is to solve for the solutions of the Hamilton-Jacobi inequality. In general, it is hard to obtain a closed form solutions due to the nonlinear nature of the inequality. In this paper, we seek for the existence conditions of quadratic type positive semidefinite solutions of Hamilton-Jacobi inequality. This is achieved by taking Taylor's series expansion of system dynamics and investigating the negative definiteness of the associated Hamilton up to fourth order. An algorithm is proposed to seek for possible solutions. The candidate of solution is firstly determined from the associated algebraic Riccati inequality. The solution is then obtained from the candidate which makes the truncated fourth order polynomial of the inequality to be locally negative definite. Existence conditions of the solution are explicitly attained for the cases of which system linearization possesses one uncontrollable zero eigenvalue and a pair of pure imaginary uncontrollable eigenvalues. An example is given to demonstrate the application to reliable control design problem.