• 한국통신학회논문지
• /
• 제27권8A호
• /
• pp.753-762
• /
• 2002

Blind equalization of transmission channel is important in communication areas and signal processing applications because it does not need training sequences, nor dose it require a priori channel information. In this paper, an adaptive blind MMSE channel equalization technique based on second-order statistics in investigated. We present an adaptive blind MMSE channel equalization using multichannel linear prediction error method for estimating cross-correlation vector. They can be implemented as RLS or LMS algorithms to recursively update the cross-correlation vector. Once cross-correlation vector is available, it can be used for MMSE channel equalization. Unlike many known subspace methods, our proposed algorithms do not require channel order estimation. Therefore, our algorithms are robust to channel order mismatch. Performance of our algorithms and comparisons with existing algorithms are shown for real measured digital microwave channel.

• Communications for Statistical Applications and Methods
• /
• 제26권4호
• /
• pp.337-346
• /
• 2019

This paper discusses a method for obtaining nonnegative estimates for variance components in a random effects model. A variance component should be positive by definition. Nevertheless, estimates of variance components are sometimes given as negative values, which is not desirable. The proposed method is based on two basic ideas. One is the identification of the orthogonal vector subspaces according to factors and the other is to ascertain the projection in each orthogonal vector subspace. Hence, an observation vector can be denoted by the sum of projections. The method suggested here always produces nonnegative estimates using projections. Hartley's synthesis is used for the calculation of expected values of quadratic forms. It also discusses how to set up a residual model for each projection.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권5호
• /
• pp.1671-1686
• /
• 2022

Face manipulation tools represented by Deepfake have threatened the security of people's biological identity information. Particularly, manipulation tools with deep learning technology have brought great challenges to Deepfake detection. There are many solutions for Deepfake detection based on traditional machine learning and advanced deep learning. However, those solutions of detectors almost have problems of poor performance when evaluated on different quality datasets. In this paper, for the sake of making high-quality Deepfake datasets, we provide a preprocessing method based on the image pixel matrix feature to eliminate similar images and the residual channel attention network (RCAN) to resize the scale of images. Significantly, we also describe a Deepfake detector named Cascaded-Hop which is based on the PixelHop++ system and the successive subspace learning (SSL) model. By feeding the preprocessed datasets, Cascaded-Hop achieves a good classification result on different manipulation types and multiple quality datasets. According to the experiment on FaceForensics++ and Celeb-DF, the AUC (area under curve) results of our proposed methods are comparable to the state-of-the-art models.

• Journal of Electrical Engineering and Technology
• /
• 제13권6호
• /
• pp.2511-2520
• /
• 2018

In this study, Multi-Radial Basis Function Support Vector Machine (Multi-RBF SVM) classifier is introduced based on a composite kernel function. In the proposed multi-RBF support vector machine classifier, the input space is divided into several local subsets considered for extremely nonlinear classification tasks. Each local subset is expressed as nonlinear classification subspace and mapped into feature space by using kernel function. The composite kernel function employs the dual RBF structure. By capturing the nonlinear distribution knowledge of local subsets, the training data is mapped into higher feature space, then Multi-SVM classifier is realized by using the composite kernel function through optimization procedure similar to conventional SVM classifier. The original training data set is partitioned by using some unsupervised learning methods such as clustering methods. In this study, three types of clustering method are considered such as Affinity propagation (AP), Hard C-Mean (HCM) and Iterative Self-Organizing Data Analysis Technique Algorithm (ISODATA). Experimental results on benchmark machine learning datasets show that the proposed method improves the classification performance efficiently.

• Smart Structures and Systems
• /
• 제11권5호
• /
• pp.533-553
• /
• 2013

In this paper, system identification of a cable-stayed bridge in Korea, the Hwamyung Bridge, is performed using vibration responses measured by a wireless sensor system. First, an acceleration based-wireless sensor system is employed for the structural health monitoring of the bridge, and wireless sensor nodes are deployed on a deck, a pylon and several selected cables. Second, modal parameters of the bridge are obtained both from measured vibration responses and finite element (FE) analysis. Frequency domain decomposition and stochastic subspace identification methods are used to obtain the modal parameters from the measured vibration responses. The FE model of the bridge is established using commercial FE software package. Third, structural properties of the bridge are updated using a modal sensitivity-based method. The updating work improves the accuracy of the FE model so that structural behaviors of the bridge can be represented better using the updated FE model. Finally, cable forces of the selected cables are also identified and compared with both design and lift-off test values.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.663-668
• /
• 2003

This paper presents a new observation system which is useful to observe the scene of the remote controlled robot vision. This system is composed of a motionless camera and head motion detector with a motion sensor. The motionless camera has a fish eye lens and is for observing a hemisphere space. The head motion detector has a motion sensor is for defining an arbitrary subspace of the hemisphere space from fish eye lens. Thus processing the angular information from the motion sensor appropriately, the direction of face is estimated. However, since the fisheye image is distorted, it is unclear image. The partial domain of a fish eye image is selected by head motion, and this is converted to perspective image. However, since this conversion enlarges the original image spatially and is based on discrete data, crevice is generated in the converted image. To solve this problem, interpolation based on an intensity of the image is performed for the crevice in the converted image (space problem). This paper provides the experimental results of the proposed observation system with the head motion detector and perspective image conversion using the proposed conversion and interpolation methods, and the adequacy and improving point of the proposed techniques are discussed.

• Smart Structures and Systems
• /
• 제10권4_5호
• /
• pp.353-373
• /
• 2012

The instrumented Canton Tower is a 610 m high-rise structure, which has been considered as a benchmark problem for structural health monitoring (SHM) research. In this paper, an improved automatic modal identification method is presented based on a natural excitation technique in conjunction with the eigensystem realization algorithm (NExT/ERA). In the proposed modal identification method, damping ratio, consistent mode indicator from observability matrices (CMI_O) and modal amplitude coherence (MAC) are used as criteria to distinguish the physically true modes from spurious modes. Enhanced frequency domain decomposition (EFDD), the data-driven stochastic subspace identification method (SSI-DATA) and the proposed method are respectively applied to extract the modal parameters of the Canton Tower under different environmental conditions. Results of modal parameter identification based on output-only measurements are presented and discussed. User-selected parameters used in those methods are suggested and discussed. Furthermore, the effect of environmental conditions on the dynamic characteristics of Canton tower is investigated.

• Smart Structures and Systems
• /
• 제4권6호
• /
• pp.779-794
• /
• 2008

Observer-based fault detection and isolation (FDI) filter design method is a model-based method. By carefully choosing the observer gain, the residual outputs can be projected onto different independent subspaces. Each subspace corresponds to the monitored structural element so that the projected residual will be nonzero when the associated structural element is damaged and zero when there is no damage. The key point of detection filter design is how to find an appropriate observer gain. This problem can be interpreted in a geometric framework and is found to be equivalent to the problem of finding a decentralized static output feedback gain. But, it is still a challenging task to find the decentralized controller by either analytical or numerical methods because its solution set is, generally, non-convex. In this paper, the concept of detection filter and iterative LMI technique for decentralized controller design are combined to develop an algorithm to compute the observer gain. It can be used to monitor structural element state: healthy or damaged. The simulation results show that the developed method can successfully identify structural damages.

• Smart Structures and Systems
• /
• 제23권5호
• /
• pp.405-420
• /
• 2019

In this study, vibration characteristics of offshore wind turbine tower (WTT) with gravity-based foundation (GBF) are identified from dynamic responses under wave-induced excitations. The following approaches are implemented to achieve the objective. Firstly, the operational modal analysis methods such as frequency domain decomposition (FDD) and stochastic subspace identification (SSI) are selected to estimate modal parameters from output-only dynamic responses. Secondly, a GBF WTT model composed of superstructure, substructure and foundation is simulated as a case study by using a structural analysis program, MIDAS FEA. Thirdly, wave pressures acting on the WTT structure are established by nonlinear regular waves which are simulated from a computational fluid software, Flow 3D. Wave-induced acceleration responses of the target structure are analyzed by applying the simulated wave pressures to the GBF WTT model. Finally, modal parameters such as natural frequencies and mode shapes are estimated from the output-only acceleration responses and compared with the results from free vibration analysis. The effect of wave height and period on modal parameter extraction is also investigated for the mode identification of the GBF WTT.

• International Journal of Computer Science & Network Security
• /
• 제22권11호
• /
• pp.241-247
• /
• 2022

Ad hoc network is self-organized network powered by battery. The reliability of virtual user identification and channel security are reduced when SNR is low due to limited user energy. In order to solve this problem, a joint virtual user identification and channel security en/decoding method is proposed in this paper. Transmitter-receiver-based virtual user identification code is generated by executing XOR operation between orthogonal address code of transmitter and pseudo random address code of receiver and encrypted by channel security code to acquire orthogonal random security sequence so as to improve channel security. In order to spread spectrum as well as improve transmission efficiency, data packet is divided into 6-bit symbols, each symbol is mapped with an orthogonal random security sequence. Subspace-based method is adopted by receiver to process received signal firstly, and then a judgment model is established to identify virtual users according to the previous processing results. Simulation results indicate that the proposed method obtains 1.6dB E_b/N₀ gains compared with reference methods when miss alarm rate reaches 10^-3.