Search

Objectives: This paper is a simulation study of dynamic observing period adjustment (DOPA) to determine the appropriate DOPA stage value that prevents the buffer overflow in various constrained application protocol (CoAP) based monitoring systems. Methods/Statistical Analysis: To determine the appropriate DOPA stage value for various CoAP-based monitoring systems, we set a variety of simulation scenarios that use different values of simulation parameters and investigate the number of successful transmissions and dropped messages of each simulation. Then, we determine the appropriate DOPA stage value of each scenario, which prevents buffer overflow and maximizes the number of successful transmissions simultaneously. Findings: The experimental simulation is conducted under various simulation scenarios. The simulation results show that the appropriate DOPA stage value increases as the number of devices or payload size increases. It also increases when the observing period of servers or clock rate decreases. The correlations between the DOPA stage value and simulation parameters are obtained through these results, which is used for determining the appropriate DOPA stage value. Improvements/Applications: Our study can be used to provide an optimal observing period for buffer overflow prevention in various monitoring systems.

A new back propagation neural network (BPNN) model is presented to construct a plasma etch process. This is accomplished by optimizing multi-parameterized neuron gradients using genetic algorithm. The technique was evaluated with experimental data collected during the etching of silicon carbide films in a NF 3 / CH 4 inductively coupled plasma. The etch process was characterized by a 2 4 full factorial experiment plus one center point. The trained model with the resulting 17 experiments was tested with the test data consisted of 16 experiments. The etch outputs to model include etch rate, profile angle, dc bias, and surface roughness. Compared to conventional BPNN models, for all etch outputs, GA-BPNN models demonstrated improvements ranging between 26% and 83%. For another smaller size of data, the improvement was more conspicuous and it ranged between 26% and 59%. These results reveal that the proposed technique can contribute to building accurate plasma models. Moreover, the technique is general in that it can be applied to any other complex plasma data.

A low pressure etching of silicon carbide is qualitatively characterized by using a neural network. To construct a predictive model, the etch process was characterized by means of a 25 full factorial experiment. Experimental factors that were varied include radio frequency (rf) source power, bias power, pressure, O2 fraction, and gap between the plasma source and wafer. An additional 15 experiments were conducted to test the appropriateness of the trained model. An optimized etch rate model has a root mean-squared error of 12.78 nm/min. Model response surface behaviors were certified by actual measurements. Several noticeable features at lower pressure etching include a lower etch rate, inverse relationship between the source power level and the dc bias, and a smaller etch rate variation with the source power. The effect of the bias power on the etch rate or dc bias was affected little by the pressure level. Etch mechanisms for the gap variations were quite different depending on the bias powers. Several etching aspects useful for plasma control were revealed.

A plasma is a key means to deposit or etch thin films in manufacturing integrated circuits. To maintain device quality and throughput, plasma faults should be diagnosed accurately and timely. A method for plasma diagnosis is presented and this consisted of wavelets, neural network, and radio frequency impedance match monitoring system. Plasma faults were simulated with the variations in process parameters. Using the monitor system, impedance matching variables were collected, including two electrical match positions and one reflected power. The collected data were subsequently filtered using discrete wavelet transformation (DWT) and continuous wavelet transformation (CWT). A backpropagation neural network (BPNN) was used to capture causal relationships between fault symptoms and root causes. For either DWT- or CWT-filtered data, fault model was constructed using single BPNN and modular BPNN, composed of multiple single BPNNs. Model performance was examined in terms of the prediction accuracy and fault sensitivity. For single network model, DWT-based model demonstrated improved fault sensitivity compared to CWT-based model. For modular network model, both DWT and CWT were comparable in identifying plasma faults. Meanwhile, modular network model yielded more improved prediction accuracy and fault sensitivity than single network model.

Due to complex dynamics, it has been extremely difficult to model high power devices. A predictive model is constructed by using a backpropagation neural network (BPNN). The BPNN was applied to predict electrical characteristics of a reduced surface field p-channel lateral double-diffused MOSFET. Drain–source currents for applied drain–source voltages were measured with a HP4156A. Prediction performance of BPNN model was optimized with variations in training factors. With respect to the reference models, the optimized models demonstrated considerably improved predictions. Model predictions were highly consistent with actual measurements. Further improvement was obtained by constructing a modular network comprising multiple BPNNs.

Refractive index of silicon nitride film deposited in a plasma enhanced chemical deposition system is modeled by using neural network. The deposition process was characterized with a full factorial experiment. Additional 12 experiments were conducted to test model appropriateness. Predicted model behaviors were in good agreement with actual measurements. Deposition mechanisms were qualitatively examined especially with respect to the pressure. Possible interactions between the pressure and other factors (SiH4, NH3, radio frequency (RF) power, and substrate temperature) were examined on the basis of SiH/NH bond ratio. The refractive index increased with increasing either SiH4 flow rate or RF power. In contrast, the refractive index decreased with increasing NH3 flow rate. Little interactions between the pressure and RF power were observed. Pressure effect on the refractive index was quite different depending on the level of SiH4 flow rate or substrate temperature. In general, increasing the pressure increased the refractive index. Meanwhile, the refractive index was insensitive to parameter variations at relatively high pressures. The most complex interaction occurred as the pressure interacted with the temperature. Useful clues to control the refractive index were revealed from the predictive model.

Uniformity of plasma etching has been conventionally examined only in terms of etch rate. A uniformity of etching profile surface is increasingly demanded to improve process quality. This is accomplished by applying a discrete wavelet transformation (DWT) to profile images obtained with a scanning electron microscopy. Applicability of DWT-based profile uniformity was evaluated with a tungsten etch experiment, conducted in an SF/sub 6/ helicon plasma. Its suitability was investigated as a function of process parameters and scale levels. Compared to a conventional metric, the wavelet-based one characterized more effectively the uniformity of profile variations. The proposed metric can be applied to any other plasma-processed surfaces.

Plasma etching was modelled by using a generalized regression neural network (GRNN). The etching process was characterized with a statistical experimental design. Three etch responses were modelled, which include two etch rates of aluminium and silica and etching profile. GRNN prediction ability was optimized as a function of training factor. Three types of models were constructed depending on the type of prepared data. Type I model corresponds to the model constructed with the original, non-classified data. Type II and III models were built for the classified data without and with the control of data interface, respectively. Compared to type I models, type II models for two etch rates demonstrated more than 25% improvement. By the control of data interface, type III models exhibited more than 15% improvement over type II models. Classification-based models in conjunction with data control thus illustrated much improved prediction of GRNN over those for non-classified models.

By controlling the proximity relative to plasma source, silicon carbide (SiC) has been etched in a C2F6 inductively coupled plasma. This was accomplished by adjusting the gap between the wafer substrate and the plasma source. Gap effects on SiC etching were extensively examined as a function of process factors as well as at particular two conditions, high source power (plasma density) and high bias power (ion bombardment). Decreasing the gap increased the etch rates for all conditions mainly due to the increased plasma density. The DC bias induced with the gap variation played a little role in affecting the etch rate. Gap effect was more enhanced as the wafer electrode was powered with relatively larger DC bias. Factor effects on the etch rates were quite different depending on the level of plasma density or ion bombardment. Decreasing the gap resulted in a microtrench at the base of the sidewall, which was little affected by the DC bias.

Accurate and fast diagnosis of a plasma equipment is crucial to maintain device yield and throughput. A methodology is presented to identify plasma faults. Particular emphasis is placed on the use of in situ diagnostic data in conjunction with a modular backpropagation neural network (BPNN). The experimental data were collected from a real-time impedance match monitor system. Two match positions were used as the signature of an anomaly in equipment plasma. Fault patterns were experimentally generated with a variation in process factors, including radio frequency source power, pressure, O 2 and Ar flow rates. A total of 30 experiments were conducted and subsequently divided into 20 training and 10 test data. Prediction accuracy and fault sensitivity were measured as a function of training factors, hidden neuron and initial weight distribution. The sensitivity was further evaluated from the standpoint of a single and modular network. Modular network consisted of four single networks, specific to variations in process factors. Adjusting initial weight distribution improved fault sensitivity in either network. Compared to single network, modular network greatly improved fault sensitivity irrespective of thresholds.

Summary form only given, as follows. Profiles of plasma etching have conventionally been characterized by approximating the slope with an angle or anisotropy. This is critically limited in that detailed variations on the profile surface are inevitably neglected. In current high density plasma etching, this becomes more serious since unexpected microfeatures such as bowing or microtrenching are frequently formed along the profile surface.

Summary form only given, as follows. Deposition of silicon nitride (SiN) film is one of the most critical processes that determine the efficiency of solar cells. Qualities of SiN film deposited by a plasma-enhanced chemical vapor deposition depend on many process parameters. Predicting film properties is very important to their optimization as well as to gain insight into underlying deposition mechanisms. For plasma-driven processes, however, it has been a difficult task to construct prediction models due to complexity within a plasma. In this study, a predictive model for a SiN PECVD process was constructed and used to understand physical deposition mechanisms. The interpretation was mainly focused on the refractive index, particularly with respect to the substrate temperature.

Qualitative models of plasma etching are essential for understanding physical etch behaviors as well as plasma control. Artificial neural networks (ANN) have been widely used in constructing predictive etch models. In most applications, the ANN prediction performance has been examined only in terms of the training factors. A technique for building a predictive model is presented here. This is accomplished by applying a neural network to classification data while optimizing the effect of the data boundary. The technique was evaluated with plasma etch data, characterized with a statistical experimental design. The etch response modeled included the etch rates for silica and aluminum (Al) as well as Al selectivity. Compared to the models for nonclassified data, the classification-based model demonstrated improvement in the prediction errors of 36.3 and 15.8% for the Al and silica etch rates, respectively, while exhibiting a poorer prediction for Al selectivity. However, by controlling the data boundary the Al selectivity model was improved significantly by about 43.8%. Thus, the classification-based modeling technique in conjunction with the control of the data boundary is effective in improving the prediction ability of neural network models.

A prediction model for surface roughness was constructed using a neural network and atomic force microscopy. The silicon carbide etch process was characterized by a 25 full factorial experiment. The experimental ranges of process parameters were 600–900W source power, 50–150W bias power, 4–16mTorr pressure, 0–80% O2 percentage, and 6–12cm gap. The model factors were optimized by means of a genetic algorithm. The optimized model had a root mean-squared error of 0.11nm. From the model, various plots were predicted while being supported by actual measurements. The dc bias induced by each process parameter was correlated to the surface roughness. Increasing the bias power increased the surface roughness. In contrast, the surface roughness decreased as the dc bias was larger than about 600V. The surface roughness was strongly correlated to the source power-induced dc bias only at low bias powers. The pressure effect was clear only as the dc bias was maintained at 480V. For the variations in the O2 percentage, ...