• Journal of Korea Society of Digital Industry and Information Management
• /
• v.16 no.2
• /
• pp.27-33
• /
• 2020

Backward compatibility is one of the key issues for radio equipment supporting IEEE 802.11, the typical wireless local area networks (WLANs) communication protocol. For a successful packet decoding with the backward compatibility, the frame format detection is a core precondition. This paper presents a novel frame format detection method based on a deep learning procedure for WLANs affiliated with IEEE 802.11. Considering that the detection performance of conventional methods is degraded mainly due to the poor performances in the symbol synchronization and/or channel estimation in low signal-to-noise-ratio environments, we propose a novel detection method based on convolutional neural network (CNN) that replaces the entire conventional detection procedures. The proposed deep learning network provides a robust detection directly from the receive data. Through extensive computer simulations performed in the multipath fading channel environments (modeled by Project IEEE 802.11 Task Group ac), the proposed method exhibits superb improvement in the frame format detection compared to the conventional method.

• Proceedings of the KSRS Conference
• /
• v.1
• /
• pp.278-281
• /
• 2006

Korea Meteorological Administration(KMA) has issued the tropical storm(typhoon) warning or advisories when it was developed to tropical storm from tropical depression and a typhoon is expected to influence the Korean peninsula and adjacent seas. Typhoon information includes current typhoon position and intensity. KMA has used the Dvorak Technique to analyze the center of typhoon and it's intensity by using available geostationary satellites' images such as GMS, GOES-9 and MTSAT-1R since 2001. The Dvorak technique is so subjective that the analysis results could be variable according to analysts. To reduce the subjective errors, QuikSCAT seawind data have been used with various analysis data including sea surface temperature from geostationary meteorological satellites, polar orbit satellites, and other observation data. On the other hand, there is an advantage of using the Subjective Dvorak Technique(SDT). SDT can get information about intensity and center of typhoon by using only infrared images of geostationary meteorology satellites. However, there has been a limitation to use the SDT on operational purpose because of lack of observation and information from polar orbit satellites such as SSM/I. Therefore, KMA has established Advanced Objective Dvorak Technique(AODT) system developed by UW/CIMSS(University of Wisconsin-Madison/Cooperative Institude for Meteorological Satellite Studies) to improve current typhoon analysis technique, and the performance has been tested since 2005. We have developed statistical relationships to correct AODT CI numbers according to the SDT CI numbers that have been presumed as truths of typhoons occurred in northwestern pacific ocean by using linear, nonlinear regressions, and neural network principal component analysis. In conclusion, the neural network nonlinear principal component analysis has fitted best to the SDT, and shown Root Mean Square Error(RMSE) 0.42 and coefficient of determination($R^2$) 0.91 by using MTSAT-1R satellite images of 2005. KMA has operated typhoon intensity analysis using SDT and AODT since 2006 and keep trying to correct CI numbers.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.2
• /
• pp.398-406
• /
• 2011

In this paper, we introduce a new architecture of PSO-based Polynomial Neural Networks (PNN) and discuss its comprehensive design methodology. The conventional PNN is based on a extended Group Method of Data Handling (GMDH) method, and utilized the polynomial order (viz. linear, quadratic, and modified quadratic) as well as the number of node inputs fixed (selected in advance by designer) at Polynomial Neurons located in each layer through a growth process of the network. Moreover it does not guarantee that the conventional PNN generated through learning results in the optimal network architecture. The PSO-based PNN results in a structurally optimized structure and comes with a higher level of flexibility that the one encountered in the conventional PNN. The PSO-based design procedure being applied at each layer of PNN leads to the selection of preferred PNs with specific local characteristics (such as the number of input variables, input variables, and the order of the polynomial) available within the PNN. In the sequel, two general optimization mechanisms of the PSO-based PNN are explored: the structural optimization is realized via PSO whereas in case of the parametric optimization we proceed with a standard least square method-based learning. To evaluate the performance of the PSO-based PNN, the model is experimented with using Gas furnace process data, and pH neutralization process data. For the characteristic analysis of the given entire data with non-linearity and the construction of efficient model, the given entire system data is partitioned into two type such as Division I(Training dataset and Testing dataset) and Division II(Training dataset, Validation dataset, and Testing dataset). A comparative analysis shows that the proposed PSO-based PNN is model with higher accuracy as well as more superb predictive capability than other intelligent models presented previously.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 2008.03a
• /
• pp.187-194
• /
• 2008

Because the types and severities of most engine faults are various and complex, it is not easy that the conventional model based fault detection approach like the GPA(Gas Path Analysis) method can monitor all engine fault conditions. Therefore this study proposed newly a diagnostic algorithm for isolating and diagnosing effectively the faulted components of the smart UAV propulsion system, which has been developed by KARI(Korea Aerospace Research Institute), using the fuzzy logic and the neural network algorithms. A precise performance model should be needed to perform the model-based diagnostics. The based engine performance model was developed using SIMULINK. For the work and mass flow matching between components of the steady-state simulation, the state-flow library was applied. The proposed steady-state performance model can simulate off-design point performance at various flight conditions and part loads, and in order to evaluate the steady-state performance model their simulation results were compared with manufacturer's performance deck data. According to comparison results, it was confirm that the steady-state model well agreed with the deck data within 3% in all flight envelop. The diagnosis procedure of the proposed diagnostic system has the following steps. Firstly after obtaining database of fault patterns through performance simulation, then secondly the diagnostic system was trained by the FFBP networks. Thirdly after analyzing the trend of the measuring parameters due to fault patterns, then fourthly faulted components were isolated using the fuzzy logic. Finally magnitudes of the detected faults were obtained by the trained neural networks. Because the detected faults have almost same as degradation values of the implanted fault pattern, it was confirmed that the proposed diagnostic system can detect well the engine faults.

• The KIPS Transactions:PartD
• /
• v.8D no.2
• /
• pp.154-165
• /
• 2001

A number of attempts to develop methods for measuring software effort have been focused on the area of software engineering and many models have also been suggested to estimate the effort of software projects. Almost all current models use algorithmic or statistical mechanisms, but the existing algorithmic effort estimation models have failed to produce accurate estimates. Furthermore, they are unable to reflect the rapidly changing technical environment of software development such as module reuse, 4GL, CASE tool, etc. In addition, these models do not consider the paradigm shift of software engineering and information systems(i.e., Object Oriented system, Client-Server architecture, Internet/Intranet based system etc.). Thus, a new approach to software effort estimation is needed. After reviewing and analyzing the problems of the current estimation models, we have developed a model and a system architecture that will improve estimation performance. In this paper, we have adopted a neural network model to overcome some drawbacks and to increase estimation performance. We will also address the efficient system architecture and estimation procedure by a similar case-based approach and finally suggest the heuristic search method to find the best estimate of target project through empirical experiments. According to our experiment with the optimally parsimonious neural network model the mean error rate was significantly reduced to 14.3%.

• Safety and Health at Work
• /
• v.3 no.2
• /
• pp.92-100
• /
• 2012

Objectives: The French National Occupational Diseases Surveillance and Prevention Network (RNV3P) is a French network of occupational disease specialists, which collects, in standardised coded reports, all cases where a physician of any specialty, referred a patient to a university occupational disease centre, to establish the relation between the disease observed and occupational exposures, independently of statutory considerations related to compensation. The objective is to compare the relevance of disproportionality measures, widely used in pharmacovigilance, for the detection of potentially new disease ${\times}$ exposure associations in RNV3P database (by analogy with the detection of potentially new health event ${\times}$ drug associations in the spontaneous reporting databases from pharmacovigilance). Methods: 2001-2009 data from RNV3P are used (81,132 observations leading to 11,627 disease ${\times}$ exposure associations). The structure of RNV3P database is compared with the ones of pharmacovigilance databases. Seven disproportionality metrics are tested and their results, notably in terms of ranking the disease ${\times}$ exposure associations, are compared. Results: RNV3P and pharmacovigilance databases showed similar structure. Frequentist methods (proportional reporting ratio [PRR], reporting odds ratio [ROR]) and a Bayesian one (known as BCPNN for "Bayesian Confidence Propagation Neural Network") show a rather similar behaviour on our data, conversely to other methods (as Poisson). Finally the PRR method was chosen, because more complex methods did not show a greater value with the RNV3P data. Accordingly, a procedure for detecting signals with PRR method, automatic triage for exclusion of associations already known, and then investigating these signals is suggested. Conclusion: This procedure may be seen as a first step of hypothesis generation before launching epidemiological and/or experimental studies.

• Journal of the Korean earth science society
• /
• v.32 no.6
• /
• pp.665-673
• /
• 2011

In order to estimate the hydrocarbon reserves, the porosity of the reservoir must be determined. The porosity of the area without a well is generally calculated by extrapolating the porosity logs measured at wells. However, if not only well logs but also seismic data exist on the same site, the more accurate pseudo porosity log can be obtained through artificial neural network technique by extracting the relations between the seismic data and well logs at the site. In this study, we have developed a module which creates pseudo porosity logs by using the polynomial neural network method. In order to obtain more accurate pseudo porosity logs, we selected the seismic attributes which have high correlation values in the correlation analysis between the seismic attributes and the porosity logs. Through the training procedure between selected seismic attributes and well logs, our module produces the correlation weights which can be used to generate the pseudo porosity log in the well free area. To verify the reliability and the applicability of the developed module, we have applied the module to the field data acquired from F3 Block in the North Sea and compared the results to those from the probabilistic neural network method in a commercial program. We could confirm the reliability of our module because both results showed similar trend. Moreover, since the pseudo porosity logs from polynomial neural network method are closer to the true porosity logs at the wells than those from probabilistic method, we concluded that the polynomial neural network method is effective for the data sets with insufficient wells such as F3 Block in the North Sea.

• The Journal of the Acoustical Society of Korea
• /
• v.24 no.2
• /
• pp.103-110
• /
• 2005

Providing adequate amount of spatial impression for spaciousness) has been known to be one of the most important design considerations for the good acoustics of rooms for music. and the measurement, of room acoustics using parameters. such as LEF and IACC, forms an essential part of such evaluation. However. it is unavoidable to use different transducers (figure of eight microphones. head and torso) for the measurement of each parameter and it tends to make the measurement procedure complicated. The Present work tried to provide a simpler way to measure these binaural room acoustic parameters including monaural ones with a single measurement system using both spatial information collected through a 5-channel microphone and a trained neural network. A computer simulation program, CATT-Acoustic V7.2. which allowed us to obtain exactly the same spatial information as a 5-channel microphone was used. since it requires quite a large amount of data for practical training of a neural network. Since each reflection has different energy. delay and direction, energy should be integrated properly. the concept of ray tracing method was applied inversely in this work. Also applying weightings according to the delay times was considered in this work. Finally, predicted results were compared with the measured data md their correlations were analyzed and discussed.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2004.10a
• /
• pp.433-436
• /
• 2004

In this paper, we introduce a new Fuzzy Polynomial Neural Networks (FPNNS)-like structure whose neuron is based on the Fuzzy Set-based Fuzzy Inference System (FS-FIS) and is different from that of FPNNS based on the Fuzzy relation-based Fuzzy Inference System (FR-FIS) and discuss the ability of the new FPNNS-like structure named Fuzzy Set-based Polynomial Neural Networks (FSPNN). The premise parts of their fuzzy rules are not identical, while the consequent parts of the both Networks (such as FPNN and FSPNN) are identical. This difference results from the angle of a viewpoint of partition of input space of system. In other word, from a point of view of FS-FIS, the input variables are mutually independent under input space of system, while from a viewpoint of FR-FIS they are related each other. The proposed design procedure for networks architecture involves the selection of appropriate nodes with specific local characteristics such as the number of input variables, the order of the polynomial that is constant, linear, quadratic, or modified quadratic functions being viewed as the consequent part of fuzzy rules, and a collection of the specific subset of input variables. On the parameter optimization phase, we adopt Information Granulation (IC) based on HCM clustering algorithm and a standard least square method-based learning. Through the consecutive process of such structural and parametric optimization, an optimized and flexible fuzzy neural network is generated in a dynamic fashion. To evaluate the performance of the genetically optimized FSPNN (gFSPNN), the model is experimented with using the time series dataset of gas furnace process.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.28 no.3
• /
• pp.31-47
• /
• 2003

Quality design in practice highly depends on human designer's intuition and past experiences due to lack of formal knowledge about the relationship among 10 variables. This paper represents an data mining approach for developing quality design support system that integrates Case Based Reasoning (CBR) and Artificial Neural Networks (ANN) to effectively support all the steps in quality design process. CBR stores design cases in a systematic way and retrieve them quickly and accurately. ANN predicts the resulting quality attributes of design alternatives that are generated from CBR's adaptation process. When the predicted attributes fail to meet the target values, quality design simulation starts to further adapt the alternatives to the customer's new orders. To implement the quality design simulation, this paper suggests (1) the data screening method based on ξ-$\delta$ Ball to obtain the robust ANN models from the large production data bases, (2) the procedure of quality design simulation using ANN and (3) model management system that helps users find the appropriate one from the ANN model base. The integration of CBR and ANN provides quality design engineers the way that produces consistent and reliable design solutions in the remarkably reduced time.