In this paper, we propose a new multiple access scheme for VSAT(Very Small Aperture Terminal) data communications. The proposed scheme has better performance in terms of delay throughput than Controlled Multiaccess Scheme which has excellent performance. The proposed scheme use the spare reservation method, adaptation method to offered load, and fixed bandwidth reservation method. In this paper, we analyze the performance using simple queueing model and present the simulation results. When network traffic is very low, the new scheme and the controlled multiaccess scheme have almost the same access delay characteristic. As the network load becomes higher, the new scheme's access delay gain becomes larger. Futhermore, even when the network traffic is very high it is possible to access satellite link within the delay similar to one round-trip delay. In addition to access delay performance enhancement, the new scheme has facility that supports fixed bandwidth reservation. So it shows more enhanced performance in the environment that stream traffic is dominant such as in VSAT communication environment.

A combined random/reservation multiple access protocol is proposed which can provide services for packet transmission in mobile satellite systems between mobile statons, or between mobile stations and fixed stations. Random multiple access protocol and reservation multiple access protocol which are currently employed in most satellite communication systems have some strengthes and some weeknesses in according to the kind of user and traffic. In this paper, a combined random/reservation multiple access protocol with better characteristics is proposed. The models of the modified random access protocol and the proposed access protocol is setted and analyzed. The performance of the PDAMA protocol, the random access protocol and the proposed access protocol is compared using simulation. For small packet arrival rate, the performance of the proposed access protocol is close at that of the modified random access protocol, and better than that of the PDAMA protocol. As the packet arrival rate is increased, the modified random access protocol is saturated and unstable at 0.23, and the performance of the proposed access protocol is better than that of the PDAMA protocol.

The performance of continuous phase modulation signals is well known for the coherent demodulation. But the carrier recovery circuits of the coherent receiver have long acquisition time and the receiver experiences high error floors for fading channels. In this paper, we propose the differential demodulation of 2-h continuous phase modulation signals. The sets of modulation indices of 2-h phase codes adequate to the differential demodulation for differentially encoded input are obtained and the average bit error probability in Additive White Gaussian Noise environments is derived and analyzed.

In the realization of interface between subscriber and network, D-channel may deteriorate the performance of switching system in handling subscriber. Thus, D-channel must be designd in a way to interface efficiently with the system while maintaining exciting performance. This objective can be obtained by through consideration of the structure of the system. This paper shows the method to design D-channel, which has significant relation with the structure of the system whenrealizing the interface between subscriber and network, in adding ISDN function. The goal of performance is also proposed.

The propagation prediction within a cell coverage in land mobile radio service is very important. The propagation loss is presented in a A+B logS110TR form, where both A and B are the parameter as function of the frequency and the antenna height and R is the distance of between base and mobile station. The propagation prediction is Cheju area is not easy, because a great number of peaks are found here and there at the foot of the Hanla Mt. The characteristics of radio propagation in Cheju area are measured for the Seorum transmitter site. The formular of correction which is regard to the configuration of the ground is presented, and the predicted values are compared with the measured one.

Interpattern association (IPA) model which the interconnection weight matrix(IWM) is constructed by the association between patterns is effective in similar pattern recognitions. But, if the number of reference patterns is increased, the ability of recognition is decreased. Using a hierarchical recognition algorithm which adopts the tree search strategy, we classified reference patterns into sub-groups by similarity. In IPA model, if input includes random noise we make it converge to reference pattern by means of input includes random noise we make it converge to reference pattern by means of increasing the number of pixels of prohibited state in IWM. In relation to reference patterns the pixel of prohibited state made partially prohibited state of no connected state using which is not included common and feature regions by each reference patterns.

The Curie point annealing effects on electrophysical phenomena at the magnetized strontium ferrite(SrO$\cdot$ 6$Fe_{2}O_{3}$) ceramics electrode/10$^{-3}$M KC1 aqueous electrolyte interfaces have been studied using cyclic voltammetric, normal pulse voltammetric, chronocoulometric, and electrochemical impedance techiques. After the Curie point annealing the magnetic flux densities of the speciment was decreased from 900-1100 gauss to 1-2 gauss, i.e. demagnetized. The real component of interfacial impedance was decreased from 7280-7320 ohm to 790-830 ohm. The adsorption and the charge on the electrical double layer was increased from 0 $\mu$C to -58 $\mu$C. The Curie point annealing and the related electrical double layer effect can influence not only the electrophysical properties of the strontium ferrite ceramics electrode itself but also the electrochemical phenomena at the electrode interface. This experimental results suggest that the Curie point annealing and the related electrical double layer effect can be applied to electrochemical magnetic sensors.

A very fine pattern formation process using double exposure is investigated, which can overcome the resolution limit of optical wafer stepper. The very fine pattern can be obtained by moving the edge profile of large pattern by means of moving the stepper stage. The simulation results show that the light transmittance decrease bellow 9%, and the contrast increase to 16.6% for the 0.3$\mu$m photoresist pattern exposeed by the double exposure using i-line wafer stepper. And the experimental results show that fine photoresist pattern as short as 0.2$\mu$m can be obtained without a loss of photoresist thickness. Also, it proves that the depth of focus for 0.3$\mu$m pattern is longer than $1.5\mu$m. And, the very fine negative photoresist pattern was formmed by using the double exposure technique and the image reversal process.

The characteristics of 3 inch wafer scale GaAs epitaxial wafer grown by molecular beam epitaxy for LSI process application were studied. The thickness and doping uniformity are characterized and discussed. The growth temperature and growth rate were $600^{\circ}C$ by pyrometer, and 1 $\mu$m/h, respectively. It was found that thickness and doping uniformity were 3.97% and 4.74% respectively across the full 3 inch diameter GaAs epitaxial layer. Also, ungated MESFETs have been fabricated and saturation current measurement showed 4.5% uniformity on 3 inch, epitaxial layer, but uniformity of threshold voltage increase up to 9.2% after recess process for MESFET device.

A high-resolution transmission electron microscopy study of the solid phase crystallization of the amorphous silicon thin films, deposited on SiOS12T at 52$0^{\circ}C$ by low pressure chemical vapor deposition and annealed at 55$0^{\circ}C$ in a dry N$_{2}$ ambient was carried out so that the arrangement of atoms in the crystalline silicon and at the amorphous/crystalline interface of the growing grains could be understood on an atomic level. Results show that circular crystalline silicon nuclei have formed and then the grains grow to an elliptical or dendritic shape. In the interior of all the grains many twins whose{111} coherent boundaries are parallel to the long axes of the grains are observed. From this result, it is concluded that the twins enhance the preferential grain growth in the <112> direction along {111} twin planes. In addition to the twins. many defect such as intrinsic stacking faults, extrinsic stacking faults, and Shockley partial dislocations, which can be formed by the errors in the stacking sequence or by the dissociation of the perfect dislocation are found in the silicon grain. But neither frank partial dislocations which can be formed by the condensation of excess silicon atoms or vacancies and can form stacking fault nor perfect dislocations which can be formed by the plastic deformation are observed. So it is concluded that most defects in the silicon grain are formed by the errors in the stacking sequence during the crystallization process of the amorphous silicon thin films.

A novel concept for the analysis of planar devices with a field limiting ring(FLR) is presented which allows analytic expressions in a normalized form for the potential distributions of FLR structure. Based on the method of image charges the main and ring junctions with identical cylindrical edges are kept to be two different equipotential surfaces. The potential relations between main and ring junction of the FLR structure are compared with 2-dimensional device simulation program. MEDICI. A good accordance is found. Comparisions with experimental data reported for the optimum ring spacing and the relative improvement of the breakdowm voltages in the FLR sturcture show the validity of the concept. The normalized expressions allow a universal application regardless to the junction depths and background doping levels.

Low cost high efficiency single crystal silicon solar cells for terrestrial applications have been fabricated by using inexpensive materials such as solar grade silicon wafer and pastes, and mass production processes such as screen printing and spray. Under 100 mW/cm$^2$ (AM 1.5) and $25^{\circ}C$ conditions conversion efficiency of 16.48% was obtained by anon fire-thru process and 15.55% by fire-thru process.

In this paper a Duty Effective Overlap Scanning method (DEOS) for the improvement of driving efficiency of LC displays which have the RMS voltage responding characteristics is proposed and new processing method of data signals for optimum application of this method is also proposed. Proposed method has a few advantages such as the increment of duty ratio the increment of driving power loaded on LC cell and the decrement of RMS voltage error rate caused by signal attenuation on electrodes composing of display when compared with the conventional method which is called as optimum voltage amplitude selection method. And also by adopting new data signal processing method which has 3 kinds of voltage levels additional advantage much improving crosstalk phenomenon which is the most serious problems of simple matrix structured display is obtained. For the characteristic estimation new mathematical representation for new overlap scanning method and data signal processing method are induced and defined. And by the defined formula and simulation the characteristics of the proposed method and the conventional method are compared and analyzed. As a result of estimation this new method being able to optimize the overlap rate of scan signal and using 3 levels of data signals has the characteristic which can improve the driving efficiency of LC displays.

In this pape, we propose a graph-theoretic approach for solving the allocation problem for the synthesis of datapaths based on multiport memories. An efficient algorithm is devised by using the weighted bipartite matching algorithm to assign variables to each port of memory modules. The proposed algorithm assigns program variables into a minimum number of multiport memory modules such that usage of memory elements and interconnection cost can be kept minimal. Experimental results show that the proposed algorithm generates the datapaths with fewer registers in memory modules and less interconnection cost for several benchmarks available from the literatures.