Catálogo de publicaciones - libros

A fast search algorithm of periodical and symmetrical modulation pattern is proposed in this paper. The algorithm is effective for the dense pulse deinterleaving of nontraditional radars. Experimental results show that the average accuracy rate of pulse deinterleaving is about 95% and the average missing rate of pulse deinterleaving is about 5% by the algorithm in dense pulse environment.

This paper presents multiple granular descriptions of a signal character and the significance of different granular analyses in signal processing. After given the concepts of granularity, we discuss the relation of different granularity, define the concepts of coarse and fine granularity and propose a granular analysis method (GAM) with which automatically choosing the suitable granularity to analyze a signal. The experimental results of extracting fine characters of a 2FSK signal show the efficiency of the method.

A novel adaptive leakage suppression method based on recurrent wavelet neural network (RWNN) in phase-coded modulation continuous wave (PCM-CW) radar is proposed in this paper. In the proposed model, the orthogonalized received signals do cross multiplication with the orthogonal local reference signals. Based on the characteristics of trigonometric function, the differencing output of the two channels effectively suppresses the leakage from transmitter while retains the interested target echoes. Considering rigorous requirements in military, blind channel equalization based on RWNN is applied to compensate the inequality in the two channels in phase property and amplitude gain to achieve suppression ratio higher than 30dB. The small size and high efficiency of RWNN make it is well suited to be utilized in the real time leakage suppression. The results of theoretical analysis and simulation on the detecting performance of the radar both show the validity and practicability of the proposed method.

Recently, Li et al. proposed a password authentication scheme based on a neural network in a multi-server environment. The scheme, however, is susceptible to off-line password guessing attacks. Accordingly, this paper will demonstrate the vulnerability of Li et al.’s scheme regarding off-line password guessing attacks, and then present an improvement to isolate such problems. The proposed scheme can withstand off-line password guessing attacks, and also provide mutual authentication.

An emulator neural network (ENN) and a parametric evaluation neural network (PENN) are constructed to facilitate a substructural parametric identification process for post-earthquake damage diagnosis of civil structures by the direct use of dynamic response measurements under base excitations. The rationality of the proposed methodology is explained, and the theory basis for the construction of two neural networks is described according to the discrete time solution of the state space equation of a substructure. An evaluation index called root-mean-square prediction difference vector (RMSPDV) is presented to evaluate the condition of a object substructure. Based on the trained ENN and PENN, the inter-storey stiffness parameters of the object substructure are identified with enough accuracy. The sensibility and the performance of the proposed methodology under different base excitations are examined using a multi-storey shear building structure by numerical simulations.

Today, gifted and talented education becomes an important part of school education. All school staff has increased awareness and knowledge about that. They develop a special program for identification of gifted student and a curriculum for them. In addition, existing gifted education pays too much attention to their curriculum, such as a curriculum compacting, acceleration, and an ability clustering. Currently, the identification of gifted student mainly depends on a simple identification test based on their age. But, the test results could not reveal the “potentially gifted” students. In this paper, we proposed a neural network model for identification of gifted student. With a specially designed questionnaire, we measure implicit capabilities of giftedness and cluster the students with similar characteristics. The neural network and data mining techniques are applied to extract a type of giftedness and their characteristics. To evaluate our model, we apply our model to the science and liberal art filed in Korea to identify gifted student and their type of giftedness.

Due to the existence of strong correlation between database metrics and response times, an online discovery quantitative models system of web service management with the linear least-squares regression algorithms was proposed. The model used the stepwise linear regression algorithms to choose a particular subset from the numerous metrics as the explanatory variables of the model, so it can be updated continuously in response to the changes made in provider configurations and the evolution of business demands. The simulation experiment for Oracle Universal Database under a TPC-W workload chose three most influential metrics that weight 66% of the variability of response time.The results show that the effectiveness of quantitative model constructing system and model constructing algorithms.

A String Graph(SG) and Alive String Graph(ASG) were defined to facilitate a static analysis of completed and counted games of Go. For the judgment of life and death, rules are applied to the situation where a stone is included and not included, and these rules are defined as a String Reduction (SR), Empty Reduction (ER), Edge Transform (ET), and Circular Graph (CG) when the stone is not included, and a Dead Enemy String Reduction (DESR) and Same Color String Reduction (SCSR) when the stone is included. Whether an SG is an ASG or not is then determined using these rules. The performance of the proposed method was tested using a problem set of games played by professional players, and all the games had been played to completion and counted. The experiment determined the error on the judgment of life and death. The test was performed on the final positions of the 20 games.  The total number of stones was 5,367 and the number of strings was 772. The experimental results produced a very low error ratio for the judgment of static life and death, where the error ratio for the stones was 0.18% and that for the strings was 1.16%.

This paper deals with research on an intelligent game character that judges the game’s physics situation and takes intelligent action in the game by applying a physics engine. The algorithm that recognizes the physics situation uses momentum back-propagation neural networks. In the experiment on physics situation recognition, a physics situation recognition algorithm where the number of input layers (number of physical parameters) and output layers (destruction value for the master car) is fixed has shown the best performance when the number of hidden layers is 3 and the learning count number is 30,000. Since we tested with rigid bodies only, we are currently studying efficient physics situation recognition for soft body objects.

A novel algorithm of document clustering based on non-negative sparse analysis is proposed. In contrast to the algorithm based on non-negative matrix factorization, our algorithm can obtain documents topics exactly by controlling the sparseness of the topic matrix and the encoding matrix explicitly. Thus, the clustering accuracy has been improved greatly. In the end, simulation results are employed to further illustrate the accuracy and efficiency of this algorithm.