A finite element method is used to analyze the vibrational characteristics of a quartz crystal tuning fork. From the information obtained by the calculation, the base length of a symmetrical tuning fork can be determined for stability of the fundamental resonant frequency and for very small stresses at the base end. Moreover, the quantitative reason why the effective resonant quality factor Qe of the supported tuning fork with arm corners cut off is highly promoted becomes clear from the analysis in this paper. This desirable characteristic is also realized in the tuning forks either with the circle hole at slit end or with the wide base portion. As for an unsymmetrical tuning fork which is frequently made in manufacturing process, the effects of dimensional unsymmetry on the resonant frequency are analyzed. For compensation of its characteristics caused by unsymmetry, the method of adjusting the length of an arm is usually adapted. It is ascertained whether or not this method is effective for high Qe and high stability of the resonant frequency against change of support condition.

A simple and practical method for arithmetic computations in a finite field GF (pn) is presented. Since number of elements of GF (pn) is pn, the method uses representation of a field element as an integer modulo pn. The method also uses log and antilog tables. Only one n-th of memories are required to store the log and antilog tables of this paper in comparison with ordinary log and antilog tables. It is not necessary to compute by using vectors. It is also shown that Zech's logarithm defined by Conway can be easily computed by using the log and antilog tables of this paper. The results of this paper are useful for decoding error-correcting cyclic codes.

Fine adjustment of wide-bandwidth SAW filter responses by trimming the interdigital transducer finger lengths was successfully achieved. In SAW devices, second-order effects, such as end effects, diffraction and bulk mode generation, disturb the simple design performances. The slight difference in wide-bandwidth frequency responses between the theoretical and experimental responses was modified by trimming the IDT finger lengths by hand.

The theoretical approximate model for packet switching multi-access system: the periodic reservation ALOHA (PR-ALOHA) is described. The performance tradeoffs of PR-ALOHA are discussed with analytic results using an approximate model obtained in this paper. By using these results, the optimum design parameters which maximize the channel utilization are given. Numerical results indicate that the PR-ALOHA offers the best throughput-delay performance at the medium throughput region.

Measurements of a far-field gain of a large Cassegrain antenna with shaped reflectors by defocusing method are discussed. Defocused-path-length variation of a shaped antenna does not considerably differ from that a conventional Cassegrain antenna. In a model antenna, even when the maximum spherical phase error is 180 degrees, defocusing method can give the far-field gain within 0.1 dB error, if the illumination does not change by defocusing.

Error correcting and/or error detecting codes have been successfully used to improve the reliability of computer memories. To improve for error control in memory systems organized to have b bit per package, a new class of linear codes for simultaneous error correction and error detection is given. We refer to a group of b bit as a byte. This paper provides a new class of binary error correcting codes to correct single bit errors and detect single byte errors (SEC-Sb ED codes), and to correct single bit errors and detect double bit errors and single byte errors (SEC-DED-Sb ED codes). Also, an improved byte error correction method which is called erasure correction method is proposed using the SEC-Sb ED codes. The decoders for the proposed codes require very small amounts of extra circuitry over that required for SEC-DED (Single Error Correcting-Double Error Detecting) codes. The decoding speed is very high-equal to that of SEC-DED codes. The proposed SEC-DED-Sb ED codes for b4 require about one more check bit and for b8 require about four more check bits than SEC-DED codes.

A novel energy relation in eigen modes of transmission line is herein proposed and the variational expressions for propagation constant of various transmission lines can be systematically derived through this relation.

The problem of concurrent access to an index structure is that of allowing a maximum number of processes to operate without impairing the correctness of their operations. Recently, Bayer et al. (1977) presented some simple locking protocols to the problem of concurrent access to indexes which are implemented B*-trees. In this paper, we give a new locking protocol that achieves higher concurrency than that of Bayer et al., by minimizing the area on which the exclusive locks are placed. We also show that this solution is deadlock free.

Optimum inspection schedule is discussed to maximize the availability for a two-unit parallel redundant system when a failure of unit is detected only by actual inspection. The stationary availability is derived by applying Piecewise Markov Process under the assumption that a failure distribution is exponential and other distributions are arbitrary.

Average intensity of far-field radiation patterns of laser light emitting from an end face of the multimode optical fiber is theoretically analyzed by assuming a Gaussian autocorrelation function for the refractive-index inhomogeneities in an optical fiber. The average intensity distribution of far-field patterns is characterized by both the statistical scale of index fluctuations and the fiber parameters of core radius and refractive-index profile. It is shown that the on-axis maximum intensity decreases as the correlation length of refractive-index inhomogeneities becomes small compared with the wavelength of light, whereas the off-axis average intensity decays gradually with the increase of the distance from the axis. The average intensity distributions obtained here are found to be nearly Gaussian for both types of the optical fiber with step-index and graded-index profiles of the refractive index.

Simple formula has been proposed to estimate baseband bandwidth of an optical transmission line made of different fibers spliced together. Using the formula, cumulative bandwidth is predicted merely from the bandwidth for each individual fiber. The usefulness of the formula has been confirmed by experiments.

Heat transfer mechanism for the compatible repeater housing, used for coaxial transmission systems, is studied. Temperature rise for the repeater housing and repeater surface is calculated in detail. An efficient improvement plan for the thermal characteristics of the new repeater housing is suggested and confirmed by the experiment.

A new practical parallel garbage collection algorithm and its correctness proof are presented. Some results of its implementation applied to a small size LISP system are reported. The algorithm, which is constructed on the condition that all active lists are bound in the linear stack, needs only one mark bit for each node and a scan-request-flag (SREQ) for the system. The minimum utilization of critical sections, where the operations on SREQ are performed, dispenses with the need for multicolored marking, such as white, gray or black. Also marking efficiency improves since it enables applying a recursive trace-and-mark method instead of a simple scan-and-markmethod. An outline of the correctness proof is as follows : It is proned by induction that the relation, wherein the active-node-set is a subset of the mark-reachable-set, which is defined as the set of all the nodes to which some links exist from any nodes already marked, is retained during the mark propagating phase for any list processing primitives, so that the algorithm guarantees that no active nodes are reclaimed. Last, the implementation of a small size LISP system with this algorithm is reported. The comparison between the parallel system and the regular one is discussed on the basis of actual data.

Dynamic properties of electrical circuits containing a Josephson junction are studied analytically by approximating the sinusoidal current-phase relation by a triangular one. Analytical expressions are obtained for the circuit-parameter dependences of transient voltage waveforms, ac ripples of stationary-state voltage oscillation and minimum currents in several typical circuits, which are shown to be in good agreement with numerical results. It is also shown that these quantities can be well characterized in simple forms in terms of a frequency peculiar to each circuit.

Some practically optimal conditions for coexistence of analog and digital systems in an intersystem interference environment may be found in a compromise between both systems. To find them, as a first step we have investigated and discussed the effects of multiple wideband FM systems upon an M-ary coherent PSK system under the assumption that both systems are sharing the same radio frequency bands. For that, in this paper we propose an analysis method which is directly applicable to the wideband FM interferers. The symbol error performance of MCPSK system in the presence of Gaussian noise and cochannel or adjacent channel interference from wideband FM systems is evaluated and given in graphes as functions of CNR, CIR and normalized carrier separation. From the results obtained here, one can expect some optimal or suitable operating conditions (frequency allocation, bandwidth and power) for a PSK channel in the intersystem interference from analog FM channels.

Linear array antennas weighted by selective withdrawal of elements have been optimized for sidelobe reduction. The optimization was carried out for antennas with total elements up to 41 by modifying implicit enumeration algorithm for 0-1 type integer optimization. Results obtained, for example, show that the sidelobe level optimized decreases almost monotonically with an increase in the total number of elements, and that the sidelobe level of about 19.0 dB is obtainable for the antenna with total elements of 41.

All of the characteristics of the batwing radiator, which is the heart of the super turnstile antenna system, are theoretical calculated by using the Moment Method.

Josephson junctions with semiconductor Te barriers have been studied. First, the junctions were fabricated with the process including oxidation of the evaporated Te film surface. The process was generally employed to avoid shorts through pinholes in Te film. Intensive studies on the influence of the Te barrier thickness and oxidation conditions on the junction characteristic parameters were made. The Te layer acts not only as the barrier for tunneling, that is the normal tunneling resistance shows exponential dependence on the Te layer thickness, but also induces proximity effect at the metal interface. The effect was found to play an important role in the junction performance. It was also found that, in the oxidation process, Te itself is oxidized and the junction performance is largely influenced by the Te oxide layer existance. In the second step, junctions free from oxide layer, which showed Josephson effect up to the Te barrier thickness of 40 nm, have been fabricated. By employing smooth surfaced Nb bottom electrode as well as the photolithographic technique comprising the trimming process, uniform junctions perfectly free from pinhole shorts have been realized.

We consider an initial value problem of the K-dV equation. The initial function is, in our case, assumed to be periodic with respect to the x-coordinate. Such a periodic function is inserted as the potential function in the associated Schrödlinger equation. The equation is, therefore, of Hill's type. A stable region in the stability chart in the λ-A plane for Hill's equation corresponds to a band which consists of continuous eigenvalues, were A is the amplitude of the periodic function and λ the eigenvalue. As the potential function, we consider a function in which dips of cosine functions of one-period stand side by with constant interval. When the interval between two consecutive cosine pulses is made infinite, the eigenvalues for one potential of cosine type is obtained. In this limiting case, clear correspondence between the numbers of eigenvalues and solitons is obtained, and soliton regions are precisely determined. This analysis can be compared with results of computer experiments on the K-dV equation. On basis of our method, it is seen that the determination of soliton region is also usable for the determination of the number of solitons.

Properties of current gain variation in n+ -p -n -n+ power transistors preceding to the thermal mode second breakdown and their applications to evaluate the resistance of second breakdown are presented. The characteristics of current gain in power transistors with junction temperature and current density are examined under high voltage operating conditions which are sufficient to generate thermal instabilities and second breakdown. Applying these properties, processes of generation of thermal instabilities and second breakdown in power transistors are investigated. Predictions of second breakdown initiation are made by monitoring the variation of the base current under the constant voltage and constant current conditions at the collector-base junction. The waveforms of the base current under these conditions show the typical changing point just before the second breakdown initiation. It is concluded that, the characteristic phenomena are related to the variation of the current gain, and the changing point corresponds to the generation of thermal instabilities. The forward biased second breakdown resistance can be measured without damage or degradation of the sample transistors, if the supplied power has been removed when the base current reaches the changing point.