In this paper, we propose a partial scan test method which can be applied to large VLSI designs. In this method, it is not necessary to hold neither scanned nor unscanned flip-flops during scan in, test application,or scan out. This test method requires almost identical design for testability modification and test wave form when compared to the full scan test method, and the method is applicable to large VLSI chips. The well known FAN algorithm has been modified to devise to sequential ATPG algorithm which is effective for the proposed test method. In addition, a partial scan algorithm which is effective for the proposed test method. In addition, a partial algorithm determined a maximal set of flip-flops which gives high fault coverage when they are unselected. The experimental resutls show that the proposed method allow as large as 20% flip-flops to remain unscanned without much decrease in the full scan fault coverage.

In this paper, we present a fast and efficient Iterative Improvement Partitioning(IIP) technique for VLSI circuits and hybrid bucket structures on its implementation. The IIP algorithms are very widely used in VLSI circuit partition due to their time efficiency. As the performance of these algorithms depends on choices of moving cell, various methods have been proposed. Specially, Cluster-Removal algorithm by S. Dutt significantly improved partition quality. We indicate the weakness of previous algorithms wjere they used a uniform method for choice of cells during for choice of cells during the improvement. To solve the problem, we propose a new IIP technique that selects the method for choice of cells according to the improvement status and present hybrid bucket structures for easy implementation. The time complexity of proposed algorithm is the same with FM method and the experimental results on ACM/SIGDA benchmark circuits show improvment up to 33-44%, 45%-50% and 10-12% in cutsize over FM, LA-3 and CLIP respectively. Also with less CUP tiem, it outperforms Paraboli and MELO represented constructive-partition methods by about 12% and 24%, respectively.

This paper presents a pipeline high-level synthesis tool for designing trace-back systolic array viterbi decoder. It consists of a dta flow graph(DFG) generator and a pipeline data path synthesis tool. First, the DFG of the vitrebi decoder is generated in the from of VHDL netlist. The inputs to the DFG generator are parameters of the convolution encoder. Next, the pipeline scheduling and allocationare performed. The synthesis tool explores the design space efficiently, synthesizes various designs which meet the given constraints, and choose the best one.

A partitioned signature file is an enhancement of the signature file that divides all the signatures into blocks in such a way that each block contains the signatures with the same key. Its directory storesall the keys as meta information for avoiding unnecessary block accesses by examming them first before the acture searching of the blocks.. Efficient directory management is very important in large databasse environments since ist size gets larger proportionally to that of the database. In this paper, we first point out the problems in the directory management methods of the previous partitioned signature files, and then present a new one solving them. OUr method offers good features in the followint three aspects: (1) suitability for large database environments, (2) adaptability to dynamic situations, and (3) storage overhead for the directory. Moreover, we can seamlessly integrate it as a subcomponent into previously-developed general-purpose storage engines. These features show that our method is applicableto signature-based access structures for the content-based retrieval in various multimedia applications such as hypermedia systems, digital library systems, multimedia document systems, multimedia mailing systems, and so on.

Multi-disk data allocation problem examined in this paper is to find a method to distribute a Binary Cartesian Product File on multiple disks to maximize parallel disk I/O accesses for partial match retrieval. This problem is known to be NP-hard, and heuristkc approaches have been applied to obtain sub-optimal solutions. Recently, efficient methods have been proposed with a restriction that the number of disks in which files are stored should be power of 2. In this paper, we propose a new disk Allocation method based on Genetic Algorithm(GA) to remove the restriction on the number of disks to be applied. Using the schema theory, we prove that our method can find a near-optimal solutionwith high probability. We compare the quality of solution derived by our method with General Disk Modulo, Binary Disk Modulo, and Error Correcting Code methods through the simulation. The simulation results show that proposed GA is superior to GDM method in all cases and provides comparable performance to the BDM method which has a restriction on the number of disks.

This paper propose a hardware architecure of multimediasysgem for integrated processing of the multimedia data such as audio and video, and describes on the design and implementation of multimedia integrated processing Unit. The unit comprises most commonly needed multimedia processing function for computer-based video conference: audio-visual datacapture, playback, compression, decompression as well as interleaving/disinterleaving of compressed audio-visual data. The proposed architecture minimizes the CPU overhead that might be caused by multimedia data processing and assures the fluent data flow among system components. Also, this unit is tested and analyzed under the computer-based video conference to confirm the multimedia unit of proposed architecture using communication protocol and application software through Ethernet and FDDI (Fiber Distributed Data Interface) networks.

This paper presented a method of extracting algorithm for Edge Multiple-Valued Decision Diagrams(EMVDD), a new data structure, from Binary Decision Diagram(BDD) which is resently using in constructing the digital logic systems based on the graph theory. We discussed the function minimization method of the n-variables multiple-valued functions and showed that the algorithm had the regularity with module by which the same blocks were made concerning about the schematic property of the proposed algorithm. We showed the EMVDD of Full Adder by module construction and verified the proposed algorithm by examples. The proposed method has the visible, schematical and regular properties.

The Error Back-Propagation(EBP) algorithm is widely applied to train a multi-layer perceptron, which is a neural network model frequently used to solve complex problems such as pattern recognition, adaptive control, and global optimization. However, the EBP is basically a gradient descent method, which may get stuck in a local minimum, leading to failure in finding the globally optimal solution. Moreover, a multi-layer perceptron suffers from locking a systematic determination of the network structure appropriate for a given problem. It is usually the case to determine the number of hidden nodes by trial and error. In this paper, we propose a new algorithm to efficiently train a multi-layer perceptron. OUr algorithm uses stochastic perturbation in the weight space to effectively escape from local minima in multi-layer perceptron learning. Stochastic perturbation probabilistically re-initializes weights associated with hidden nodes to escape a local minimum if the probabilistically re-initializes weights associated with hidden nodes to escape a local minimum if the EGP learning gets stuck to it. Addition of new hidden nodes also can be viewed asa special case of stochastic perturbation. Using stochastic perturbation we can solve the local minima problem and the network structure design in a unified way. The results of our experiments with several benchmark test problems including theparity problem, the two-spirals problem, andthe credit-screening data show that our algorithm is very efficient.

The self-excited oscillation is the phenomenon which can be observed in the systems composed of nonlinear elements. The phenomenon is of fundamental importance in nonlinear systems and, as far as the design of a nonlinear system is concerned, it should be considered along with the stability analysis. In this paper, the oscillation of a system controlled by a static nonlinear fuzzy controller is theoretically addressed. First, the describing functionof a static fuzzy controller is derived and then, based on the derived describing function, self-excited oscillation of the system controlled by a static fuzzy controller is predicted. To obtain the describing function of the static fuzzy controller, a simple struture is assumed for the fuzzy controller. Finally, computer simulation is included to show an example where the describing function given in the paper is used to predict the self-excited oscillation of a fuzzy-control system.

In face recognition based on the Karhunen-Loeve approximation, amplitudespectra of Fourier transformed facial images were used. We found taht the use of amplitude spetra gives not only the shift-invariance property but also some improvment of recognition rate. This is because the distance between the varing faces of a person compared with that between the different persons perfomed computer experiments on face recognitio with varing facial images obtained from total 55 male and 25 females. We confirmed that the use of amplitude spectra of Fourier-trnsformed facial imagesgives better recognition rate for avariety of varying facial images including shifted ones than the use of direct facial images does.