Showing papers presented at "Computer Aided Systems Theory in 2011"

Parameter meta-optimization of metaheuristic optimization algorithms

Abstract: The quality of a heuristic optimization algorithm is strongly dependent on its parameter values. Finding the optimal parameter values is a laborious task which requires expertise and knowledge about the algorithm, its parameters and the problem. This paper describes, how the optimization of parameters can be automated by using another optimization algorithm on a meta-level. To demonstrate this, a meta-optimization problem which is algorithm independent and allows any kind of algorithm on the meta- and base-level is implemented for the open source optimization environment HeuristicLab. Experimental results of the optimization of a genetic algorithm for different sets of base-level problems with different complexities are shown.

Prostheses control with combined near-infrared and myoelectric signals

Abstract: Over the last decade, technical innovations have resulted in major achievements in the field of upper limb prostheses. However, the limitation of such devices is not the electro-mechanical realization, but rather the lack of astute cutaneous control-sources. This contribution demonstrates real-time detection of muscle exertions by combining myoelectric and near-infrared signals. The presented sensor technology and classification scheme have been developed for five-finger hand prostheses but can be employed to control a variety of prosthetic devices. This control mechanism only requires a patient to individually contract extant muscles, for which a minimal-threshold myoelectric or reflected near-infrared signal can be measured on the surface of the skin. Our experimental data show that the combination of both sensor types provides better classification results and surpasses the spatial resolution of a single pickup device. Features extracted from these signals can be used as input data for our existing classifier and allow compensation of muscle fatigue effects.

U-V disparity analysis in urban environments

Abstract: Traditionally obstacles detection is a great topic in computer vision applied to robotics navigation or advance driver assistance system (ADAS). Although other technologies, such as laser, obtain good results to detect obstacles in different environments, stereo vision has the advantage of providing 3D information, improving the knowledge of the environment. A study of the implementation of the u-v disparity in urban environments is presented in this paper, where several tests have been done in different situations which may be difficult to interpret by using a straightforward analysis of the u-v disparity in order to model the environment.

Rational function systems in ECG processing

Abstract: In this paper we propose to use rational function systems for processing ECG signals. In several respects our approach has advantages over the previously used transformation methods. For instance the system is designed particularly for ECG signals in the sense that the shape of the individual terms correspond to the natural segmentation of the signal. Moreover, our method is more flexible. This means that not only the coefficients but also the system itself can be optimized even from heartbeats to heartbeats. Another property worth to emphasize is the simplicity of the system. The terms are simple rational functions that can be generated by a few number of complex arithmetic operations, i.e. they are easy to work with on computers.

Precise segmentation of the optic disc in retinal fundus images

Abstract: This paper presents a methodology for the precise extraction of the optic disc in retinal images. We propose a methodology that combines an enhancement algorithm, a fast location methodology and, finally, a segmentation technique to fit the optic disc boundaries. At each stage, we analyze several techniques and we present the results obtained in a public image data set.

Improvements of the construction of exact minimal covers of boolean functions

Abstract: The calculation of an exact minimal cover of a Boolean function is an NP-complete problem. In this paper we introduce the definition of this problem and its basic solution. By using a slightly modified algorithm, we get a speed-up factor of more than 104. The main contributions of this paper are the description of an alternative approach mentioned in [15], and a remarkable improvement of this algorithm. In both cases operations of the XBOOLE library are used. Using the newly suggested algorithm, the time required for the calculation could be reduced by a factor of more than 8∗108 in comparison with the previous algorithm.

Mobile platform architecture review: android, iphone, qt

Abstract: The recent proliferation of mobile platforms makes it difficult for developers to find the most appropriate platform for one's needs or even target multiple ones. A review of key concepts on Android, iPhone and Qt points out important commonalities and differences that will help to better understand the respective platform characteristics. As mobility is an important aspect of such phones, the ability to access mobile-specific functionality is investigated. Implications at every concept visited will then point out things to keep in mind as a developer.

Analysis of selected evolutionary algorithms in feature selection and parameter optimization for data based tumor marker modeling

Abstract: In this paper we report on the use of evolutionary algorithms for optimizing the identification of classification models for selected tumor markers. Our goal is to identify mathematical models that can be used for classifying tumor marker values as normal or as elevated; evolutionary algorithms are used for optimizing the parameters for learning classification models. The sets of variables used as well as the parameter settings for concrete modeling methods are optimized using evolution strategies and genetic algorithms. The performance of these algorithms is analyzed as well as the population diversity progress. In the empirical part of this paper we document modeling results achieved for tumor markers CA 125 and CYFRA using a medical data base provided by the Central Laboratory of the General Hospital Linz; empirical tests are executed using HeuristicLab.

Choice of directions for the approximation of reachable sets for hybrid systems

Abstract: In this paper we propose an approach to over-approximate the reachable set (with bounded time and number of transitions) of a hybrid system by a finite set of polytopes. The constraints of the polytope are determined by a direction choice method. For the hybrid systems whose (1) continuous dynamics are linear, (2) invariants and guards are defined by linear inequalities, and (3) variable resets are expressed by invertible affine maps, we show that the over-approximations can be computed in polynomial time, and the overestimation can be arbitrarily reduced by decreasing the discretization time step if the continuous dynamics are all deterministic. Some experimental results are also presented to show the effectiveness of our approach.

Systolic optimization on GPU platforms

Abstract: The article presents a systolic algorithm implemented using NVIDIA's Compute Unified Device Architecture (CUDA). The algorithm works as a general disposition of the elements in a mesh by sinchronously computing basic solutions among processing elements. We have used instances of the Subset Sum Problem for evaluating to study the behavior of the proposed model. The experimental results show that the approach is very efficient, especially for large problem instances and consumes shorter times compared to other algorithms like parallel Genetic Algorithms and Random Search.

Friends radar : towards a private P2P location sharing platform

Abstract: In this paper, we propose a new approach to live location sharing among a group of friends that does not rely on a central server as a hub for exchanging locations messages, but is done in a peer-to-peer manner by pushing selective location updates via XMPP messages. That is, only those contacts specifically authorized by users will be notified of location changes via push messages, but no third-party service is involved in processing these locations. We describe an initial implementation called Friends Radar, and discuss the associated trade-offs between privacy and resource consumption.

Models and techniques for computer aided surgical training

Abstract: Laparoscopic surgery, when performed by a well-trained surgeon, is a remarkably effective procedure that minimizes complications associated with large incisions, operative blood loss and post-operative pain. It also reduces the recovery time. However, the procedure is more challenging than a conventional surgery due to the restricted vision, hand-eye coordination problems, limited working space, and lack of tactile sensation. These issues make the laparoscopic surgery a more difficult technique for medical students and residents to master.

An easy to use infrastructure for building static analysis tools

Abstract: This paper deals with design and implementation of an easy to use infrastructure for building static analyzers. The infrastructure provides an abstraction layer called a Code Listener over existing source code parsers like, for example, GCC or Sparse. It is distributed as a C++ library that can be used to build static analyzers in the form of GCC plug-ins. The interface exposed to analyzers is, however, completely independent of GCC, which allows one to run the same analyzer on top of different code parsers without a need to change anything in the analyzer. We describe the key design principles of the infrastructure and briefly introduce its application programming interface that is available to analyzers. The infrastructure is already used in research prototypes Predator and Forester, implementing advanced shape analyses, intended to operate on real industrial code.

Automatic tuning methods for MPC environments

Abstract: Model predictive control is a powerful method for controlling multivariable systems, in particular when constraints have to be taken into account. However, MPC also has negative attributes, like a high computational effort but also a non intuitive tuning. While the first issue can be addressed by use of numerically efficient optimizers, the non intuitive tuning still remains. To this end, an approach for efficient tuning of a MPC environment consisting of controller and state observer is proposed. The idea is to provide an automatic tuning strategy, such that even an unexperienced user can design a satisfactory controller within reasonable time. The proposed tuning of the state observer is done by a combination of multi model and adaptive estimation methods and for the weight tuning of the MPC objective function an additional optimization loop is applied which also accounts for the numerical condition. Finally an example is presented, where the proposed strategies were used to tune the MPC for controlling a pipeline compressor natural gas engine in a nonlinear simulation environment, yielding promising results.

A uniform classification of common concurrency errors

Abstract: Nowadays, multi-threaded programs are quite common and so are concurrency-related errors. Many works devoted to detection of concurrency errors have been published in recent years, and many of them presented definitions of concurrency errors that the proposed algorithms are able to handle. These definitions are usually expressed in different terms suitable for a description of the particular considered algorithms, and they surprisingly often differ from each other in the meaning they assign to particular errors. To help understanding the errors and developing techniques for detecting them, this paper strives to provide a uniform taxonomy of concurrency errors common in current programs, with a stress on those written in Java, together with a brief overview of techniques so far proposed for detecting such errors.

Comprehensive and automatic fitness landscape analysis using heuristiclab

Abstract: Many different techniques have been developed for fitness landscape analysis. We present a consolidated and uniform implementation inside the heuristic optimization platform HeuristicLab. On top of these analysis methods a new approach to empirical measurement of isotropy is presented that can be used to extend existing methods. Results are shown using the existing implementation within HeuristicLab 3.3.

Usability evaluation framework: automated interface analysis for android applications

Abstract: Usability evaluation of mobile applications can be a time-consuming task, especially due to the fact that field studies are often more suitable than laboratory evaluations. However, good user interface design can make the difference between product acceptance and rejection in the marketplace, and is therefore of high relevance for application developers. In this paper, we present an innovative approach for the automated evaluation of the usability of a mobile application, which is based on the background observation and analysis of a user's behavior when using the application under real-life conditions. We present first results of our ongoing research work on a usability evaluation framework, which should help developers to make mobile applications easier to learn, easier to use and aesthetically more pleasing, and thus lead to a higher usability and acceptance of the applications.

Lightweight authentication for RFID used in VANETs

Abstract: This work addresses the critical problem of authentication in RFID. It describes a new lightweight scheme for mutual authentication between readers and tags that fulfills the EPCGen2 standard and all practical requirements of low-cost RFID such as resource limitation of tags and minimal interaction between tags and readers. Furthermore, the proposal does not rely on RFID readers as they are portable, and instead of that, it bases its security on trust in the back-end server because all shared secrets are stored only by the tag and the back-end server, with no possible access by the reader at any time.

ACO-GRASP-VNS metaheuristic for VRP with fuzzy windows time constraints

Abstract: In this article we propose a methodological approach based on Soft Computing to obtain solutions to the vehicle routing problem when time windows constraints are imprecise and flexible A fuzzy model and method is obtained by applying a Fuzzy Optimization approach A new hybrid algorithm is presented and applied to real problem instances of a distribution company that combines Ant Colony Optimization, Variable Neighbourhood Search and Greedy Randomize Adaptive Search Procedure for the corresponding fuzzy optimization problem

Dynamic cellular automata-based s-boxes

Abstract: The most important elements of many block ciphers are nonlinear functions known as substitution boxes (S-boxes). Classical S-boxes are usually represented by numerical tables, which are used today in current cryptographic standards, such as Data Encryption Standard (DES) or Advanced Encryption Standard (AES), but in the result of developing methods of cryptanalysis they do not ensure enough safety of ciphers. Therefore, the open research issue now is to design new more sophisticated classes of S-boxes, in particular dynamic ones. In this paper we propose a methodology to design dynamic cellular automata (CA)-based S-boxes, which can be considered as a generator of CA-based S-boxes. We provide an exhaustive experimental analysis of the proposed CA-based S-boxes in terms of non-linearity, autocorrelation, balance and strict avalanche criterion. We show that the proposed S-boxes have high quality cryptographic properties (high non-linearity and balance, also low autocorrelation and distance to fulfill strict avalanche criterion). The interesting feature of the proposed S-boxes is a dynamic flexible structure, fully functionally realized by CA, while the classical S-boxes are represented by predefined unchangeable table structures.

Simulation-Based fitness landscape analysis and optimisation for vehicle scheduling problem

Abstract: The paper presents simulation optimisation methodology and tools for the vehicle scheduling problem (VSP) with time windows The optimisation problem statement is given The fitness landscape analysis is used to evaluate the hardness of the problem The tool for fitness landscape analysis is build up To evaluate fitness of solutions the vehicle schedule simulation model in AnyLogic 6 is developed, and Java applications generate landscape path solutions and analyse their fitness series A genetic algorithm is applied for simulation-based vehicle schedule optimisation The results of the experimental study are described

Neuronal data analysis based on the empirical cumulative entropy

Abstract: We propose the empirical cumulative entropy as a variability measure suitable to describe the information content in neuronal firing data. Some useful characteristics and an application to a real dataset are also discussed.

Particle swarm optimization with two swarms for the discrete (r

Abstract: The (r|p)-centroid problem is a competitive location problem that consists of determining optimal strategies for two competing firms, the leader and the follower, which make decisions sequentially. We propose a particle swarm optimization procedure with two swarms to solve the discrete (r|p)-centroid problem. The proposed algorithm considers a swarm for each player in the location game.

Solving the two-dimensional bin-packing problem with variable bin sizes by greedy randomized adaptive search procedures and variable neighborhood search

Abstract: In this work we present new metaheuristic algorithms to a special variant of the two-dimensional bin-packing, or cutting-stock problem, where a given set of rectangular items (demand) must be produced out of heterogeneous stock items (bins). The items can optionally be rotated, guillotine-cuttable and free layouts are considered. The proposed algorithms use various packing-heuristics which are embedded in a greedy randomized adaptive search procedure (GRASP) and variable neighborhood search (VNS) framework. Our results for existing benchmark-instances show the superior performance of our algorithms, in particular the VNS, with respect to previous approaches.

Stereo-Vision algorithm based on bio-inspired silicon retinas for implementation in hardware

Abstract: In this paper, we propose a silicon-retina-based stereo vision system, used for pre-crash warning respectively side-impact detection applications in vehicles The bio-inspired Silicon Retina sensor is a new kind of sensor, with a high temporal resolution of 1ms and a dynamic range of approx 120dB These type of imagers deliver data asynchronously and only when the intensity of the ambient light changes Therefore, the amount of data that must be processed decreases significantly compared to standard CMOS or CCD imagers The sensor uses an address-event representation (AER) protocol to transfer the event-triggered information Concerning these special output characteristics, a novel approach regarding acquisition, storage, and stereo matching of the data were implemented The concept of the algorithm is specifically targeted and optimized for an implementation in hardware, eg on a Field Programmable Gate Array (FPGA)

An evolutionary algorithm with solution archive for the generalized minimum spanning tree problem

Abstract: We propose a concept of enhancing an evolutionary algorithm (EA) with a complete solution archive that stores evaluated solutions during the optimization in a trie in order to detect duplicates and to efficiently convert them into yet unconsidered solutions. As an application we consider the generalized minimum spanning tree problem where we are given a graph with nodes partitioned into clusters and exactly one node from each cluster must be connected. For this problem there exist two compact solution representations that can be efficiently decoded, and we use them jointly in our EA. The solution archive contains two tries --- each is based on one representation, respectively. We show that these two tries complement each other well. Test results on TSPlib instances document the strength of this concept and that it can match up with the leading state-of-the-art metaheuristic approaches from the literature.

Optimized filter design for a filter bank based blocker detection concept for LTE systems

Abstract: For mobile communication systems power efficiency is a very important issue. Especially for mobile user equipments a careful management and efficient use of the limited energy ressources is mandatory. In today's user equipments quite an amount of energy is wasted. The reason for this is, that analog and digital frontend in communication systems are engineered for extracting the wanted signal from a spectral environment, which is defined in the corresponding communication standards with strict requirements. In a real receiving process those requirements can typically be considered as less critical. Sensing the environmental transmission conditions and adapting the receiver architecture to the actual needs allows to save energy during the receiving process. An efficient architecture being able to fulfill this task for a typical Long Term Evolution scenario has been disussed recently. For the implementation of this architecture, highly efficient filter approaches had to be investigated. This paper gives an overview on the basic properties of those approaches and compares it to well known filter types.

On memory capacity to implement logic functions

Abstract: The paper presents logic synthesis method targeted at FPGA architectures with specialized embedded memory blocks (EMB). Existing tools/compilers treat ROM modules described in HDLs as indivisible entities and in consequence do not ensure effective utilization of the possibilities provided by such modules. In order to address this problem effectively we propose applying functional decomposition. The main contribution is based on the use of r-admissibility measure to guide decomposition structures for ROM-based synthesis.

Application of symbolic regression on blast furnace and temper mill datasets

Abstract: This work concentrates on three different modifications of a genetic programming system for symbolic regression analysis. The coefficient of correlation R2 is used as fitness function instead of the mean squared error and offspring selection is used to ensure a steady improvement of the achieved solutions. Additionally, as the fitness evaluation consumes most of the execution time, the generated solutions are only evaluated on parts of the training data to speed up the whole algorithm. These three algorithmic adaptations are incorporated in the symbolic regression algorithm and their impact is tested on two real world datasets describing a blast furnace and a temper mill process. The effect on the achieved solution quality as well as on the produced models are compared to results generated by a symbolic regression algorithm without the mentioned modifications and the benefits are highlighted.

Pseudorandom generator to strengthen cooperation in VANETs

Abstract: A new secure communication system is here presented for spontaneous and self-managed vehicular ad-hoc networks without any infrastructure on the road or vehicles. Our proposal prevents passive behavior of users who try to take advantage of the network without cooperating in its operation. In this paper we propose the use of encrypted exchange of data as a method to strengthen cooperation in VANETs. In particular, we describe a new pseudorandom generator based on a linear feedback shift register for encrypting sent information through a stream cipher, and analyze its output statistically.