In many packet switched networks, packet sizes are basically 128 or 256 bytes, probably due to their adaptation to interactive communications with short delay requirement. However, in order to deal with message communications also, characterized as bulk, unidirectional and delay tolerable, the networks may be required to handle a larger packet size suitable for the message communications characteristics. This paper calls packet switching systems, which handle various sizes of packets, heterogeneous packet switching systems. Focusing on packet transfer cost in heterogeneous packet switching systems, this paper discusses an optimum packet size for message communications. For an optimum packet size, an approximate and analytical solution is obtained when packet header length is ignored. Moreover, strict and numerical calculation results under a general condition are also shown. It is shown that the approximation gives a smaller size than the strict solution. Further, the effect of cost reduction caused by an optimum packet size for message communications is evaluated. As a result, the cost in heterogeneous packet switching is confirmed to be reduced to about half, especially when message communications traffic is dominant.

A speaker independent algorithm is given which automatically detects the most steady-state portion of a vowel (vowel center) from continuous speech. The algorithm first extracts the segments each of which contains a vowel and, if present, pre- and/or post-vocalic liquids and semivowels, and then locates the most steady-state portion of the segment. An advantage of the algorithm is its ability to distinguish the nasal and the intervocalic liquid and semivowel segments without relying upon the formant frequencies which have been used in most of the previous work of vowel segment detection procedure. This results in a computationally simple algorithm. The test on ten sentences spoken by each of two males and two females resulted in score of 93.2% correct vowel center localization.

This paper investigates the linearity performance of analog video transmission systems which use semiconductor laser diodes. The main origins of system linearity degradation are shown to be reflected laser light and speckle noise. It is demonstrated that a laser with an external reflector produces a nonlinear response which degrades the laser modulation properties. It is also revealed that in analog video transmission systems using graded-index multimode fibers, speckle noise is so serious that a reduction in laser coherency is essential, and a new technique employing an optical feedback scheme to reduce speckle noise in investigated. The high quality of results obtained from transmission tests employing the proposed method have confirmed the feasibility of analog video transmission using semiconductor laser diodes.

Use of parity-based codes in computer systems can provide a cost-effective error detection and correction. Excellent parity-based main memory codes, such as odd-weight-column SEC-DED (Single bit Error Correcting Double bit Error Detecting) codes, can be used to arithmetic logic unit (ALU) by the technique proposed in this paper. They are more economical and better suited for a byte-sliced design of fault-tolerant computer systems than residue codes and triplication organization.

As the structural information of characters and figures, features based on closed curves have been used. However, studies of feature extraction methods have been made mainly for a single closed curve. In this paper, a fundamental computer processing system is reported for dealing with very many collections of multiple closed curves which are represented by contour lines of a topographical map. Here, we describe an algorithm for obtaining every string of chain-codes which is local information of a collection of closed curves and a tree structure which is global information of the collection through one directional and a single row-by-row scan. Finally, the results obtained from the application of the proposed algorithm to simple recognition problems of mountains are described.

A graphic assembler language (GAL) for dealing with pictures is described with examples of its applications to computer graphics. Instruction codes of GAL are composed of four arithmetic operations of polynomials, control and input-output operations, and a sentence is defined for taking the coordinates of figures in a computer.

We carried out a research for the responses of Pb thin film microbridge Josephson device fabricated by electron lithography under microwave and millimeter wave radiations. The experimental results are reported in this paper.

In this paper, it is presented that the experimental results of the magnetic bubble propagation in dual conductor current access test circuits, where the bubble-driving fields can be generated by applying the bipolar pulse currents to the apertured conducting sheets. Under the driving conditions that the current density for the first conducting sheet (I₁) was at 0.5 mA/µm, for the second conducting sheet (I₂) was at 1 mA/µm, and the frequency of these currents was at 200 kHz, the bias margin for the propagation normal to the currents flow was obtained as much as 5 Oe.

Microwave modulation in frequency region of 5.9 to 10 GHz was determined in TJS and BH lasers. Incident microwave power was less than 32 mW at 10 GHz. Modulation mechanism was regarded to be caused by driving of the injected-carrier-dencity even at such high frequency.