We present an approach to high-level synthesis for a specific target architecture-partitioned bus architecture. In this approach, we have specific goals of minimizing data transfer length and number of buses in addition to common synthesis goals such as minimizing number of control steps and satisfying given resource constraint. Minimizing data transfer length and number of buses can be very important design goals in the era of deep submicron technology in which interconnection delay and area dominate total delay and area of the chip to be designed. in partitioned bus architecture, to get optimal solution satisfying all the goals, partitioning of operation nodes among segments and ordering of segments as well as scheduling and allocation/binding must be considered concurrently. Those additional goals may impose much more complexity on the existing high-level synthesis problem. To cope with this increased complexity and get reasonable results, we have employed two ideas in ur synthesis approach-extension of the target architecture to alleviate bus requirement for data transfer and adoption of genetic algorithm as a principal methodology for design space exploration. Experimental results show that our approach is a promising high-level synthesis mehtodology for partitioned bus architecture.

This paper presents the design and implementation of an efficient algorithm for detecting untestable paths in multi-level circuits. Transforming multi-level circuit into a multiplexor-based one through BDD(binary decision diagram)construction, the proposed algorithm detects untestable paths in the transformed circuits. By constructing ENF (equivalent normal form) only for reconvergent paths, the proposed system detects and removes untestable paths efficiently in terms of the run-time and memory usage. Experimental results for MCNC/ISCAS benchmark circuits show that the system efficiently detects and removes untestable paths. The run-time and memory usage have been reduced by 37.7% and 60/9%, respectively, comapred to the previous methods.

In tis paper, sthe new data model for the hardware implementation of lempel-ziv compression algorithm was proposed. Traditional model generates the codeword which consists of 3 bytes, the last symbol, the position and the matched length. MSB (most significant bit) of the last symbol is the comparession flag and the remaining seven bits represent the character. We confined the value of the matched length to 128 instead of 256, which can be coded with seven bits only. In the proposed model, the codeword consists of 2 bytes, the merged symbol and the position. MSB of the merged symbol is the comression flag. The remaining seven bits represent the character or the matched length according to the value of the compression flag. The proposed model reduces the compression ratio by 5% compared with the traditional model. The proposed model can be adopted to the existing hardware architectures. The incremental factors of the compression ratio are also analyzed in this paper.

Human observers perceive a contour where no line or edge is given (subject contour), such as the border between gratings of thin lines offset by half a cycle. Subjective contours occur in a wide variety of circumstances in nature. A common factor in all such figures in the perception of a surface(contour) occuluding part of a background, i.e. subject contours are always accompanied by subject surfaces. Some examples about these kinds of figures are given by kanizsa trangle or square. Ph;ysiological observations are reported in which the perceptual phenomenon that some simple examples of subject contours are perceived as images (gigures; contours) through physiological experiments was investgated. In this paper, we proposed a neural network extracting the induce dstimuli for the perception of subject contours based on the visual mechnism. This visual mechnism is proposed by hubel & wiesel's paper.