Catálogo de publicaciones - libros

We propose a novel approach to chaotically mask traffic pattern according to a predetermined chaotic mode, aiming at making the masked traffic lose values for intruders to analyze for malicious activities in a mission critical network.

It proposes a chaotic secure communication scheme by introducing a concept of key function (KF) and key initial conditions (KICs). By using KF, secret signal is combined with complex chaotic signal in order to enhance the security of transmitted signal. KICs are used to enlarge minor mismatch of drive signal so as to increase initial conditions sensitivity of the auxiliary driver system. Experiment results show that proposed cryptosystems are very sensitive to KICs’ and KFs’ mismatch.

This paper addresses a special type of graph, the -neighbourhood graph, for the usage in huge multi agent systems. It can be used to establish slim communication structures in extensive groups of agents as they are present e.g. in swarm applications. We will prove some properties of -neighbourhood graphs in two- and three-dimensional Euclidean space, i.e. we show that the maximum number of incoming connections per agent is limited to a value independent from the overall number of agents in the system. For the two-dimensional case we prove a maximum in-degree of 6 ·. Furthermore, for agents interacting in three dimensions an upper and a lower bound for this value is presented.

There are restrictions of complete rationality and information symmetry, which is difficult to work in reality, in the traditional game theory. However, the evolutionary game theory, based on bounded rationality, can compensate these defects. In this paper, we set up an asymmetric model of the evolutionary game between a supplier and a retailer with asymmetric information. The evolutionary stable strategies and their premises were obtained with the replicator dynamics mechanism. The analysis in this paper has significance of explanation and direction for the supply chain management.

An optimal public investment model with two objective functions considering efficiency & equity and several constraints such as taxes and capital transfer loss are established by dividing public & private sectors and relaxing several original hypotheses respectively. And the objective functions and constraints are handled to adapt the model into the double-objective multi-constraint programming model suitable for genetic algorithm-based solution. Then encoding and decoding approaches are designed. Finally a case study is carried out to validate the proposed model and the GA-based solution.

This paper describes the application of a multi-population genetic algorithm to the selection of feature subsets for classification problems. The multi-population genetic algorithm based on the independent evolution of different subpopulations is to prevent premature convergence of each subpopulation by migration. Experimental results with UCI standard data sets show that multi-population genetic algorithm outperforms simple genetic algorithm.

Falls in the elderly people often cause serious physical injury, result in fracture, cerebral haemorrhage, even death. To find falls as earlier as possible is very important to rescue the subjects and facilitate the rehabilitation in the future. In this paper, we use a wearable tri-axial accelerometer to monitor the movement parameters of human body, and propose a novel fall detection algorithm based on non-negative matrix factorization (NMF). The input vectors are the acceleration sequences of the transverse section and the vertical axial of human body, and these vectors are decomposed via NMF. And then, a k-nearest neighbor method is applied to determine whether a fall occurred. The results show that this method can detect the falls effectively.

Content-based video indexing and retrieval algorithms are presented in this paper that aim at temporally indexing a video sequence according to actors. Our system splits a video into a sequence of a few representative frames. We use color information and then SGLD matrix on the representative frames for face region detection. Detected faces are used to build a face database. We construct eigen faces applying PCA on the faces in the face database for extracting important features. Extracted features are then used in MPM for identifying the input face from the training faces. Experimental result shows that our approach can correctly recognize 95.3% and 90.84% of the faces from the AT&T face database and video sequence respectively.

Inspired by complementary strategies, a new fault diagnostic method, which integrates with the Adaptive Resonance Theory (ART) and Artificial Immune Network (AIN), is proposed in this paper. With the help of clustering of ART neural network, the vaccines that image features of data set are extracted effectively, and then an AIN named aiNet is adopted to realize data compression. Finally the memory antibodies optimized by aiNet can be used to recognize each feature of original dataset and to realize fault diagnosis. The experimental results show that the approach is useful and efficient for the fault diagnosis of the multilevel reciprocating compressor.

This paper presents a novel approach for online multi-strokes composite sketchy shape recognition based on Bayesian Networks. By means of the definition of a double-level Bayesian networks, a classifier is designed to model the intrinsic temporal orders among the strokes effectively, where a sketchy shape is modeled with the relationships not only between a stroke and its neighbouring strokes, but also between a stroke and all of its subsequence.. The drawing-style tree is then adopted to capture the users’ accustomed drawing styles and simplify the training and recognition of Bayesian network classifier. The experiments prove both effectiveness and efficiency of the proposed method.