Search

A Wireless Sensor Network (WSN) is a wireless network that monitors physical environment conditions through resource-constrained sensor nodes and delivers data to a sink node through the network. One of the most important constraints on sensor nodes is their limited power source, which consists of small and irreplaceable batteries. Energy conservation is thus a dominant factor in WSN. Therefore, when designing a routing protocol for WSNs, it is necessary to consider the energy constraint of sensor nodes. In this paper, we consider the energy constraint of sensor nodes and propose an Energy Balancing Routing Protocol using reinforcement learning. The performance of the proposed protocol is compared to other existing energy-efficient routing protocols and the results show that the proposed protocol performs better with regards to energy saving and network lifetime.

Since Markowitz [13] propose the mean-variance efficient portfolio selection method it has been one of the frequently used approach to the portfolio optimization problem. However, as we know, this approach has critical draw backs such as unstable assets weights and poor forecasting performance due to the estimation error. In this study, we propose an improved portfolio selection rules using various distortion functions. Our approach can make up for the pessimism of economic agents which is important for decision making. We illustrate the procedure by four well-known datasets. We also evaluate the performance of proposed and many other portfolio strategies to compare the in- and out-of-sample value at risk, conditional value at risk and Sharpe ratio. Empirical studies show that the proposed portfolio strategy outperforms many other strategies for most of evaluation measures.

License plate recognition is the most promising application in IoT camera. The existing recognition algorithm shows good performance when it is used in the constrained situation such as at the entrance to the parking lot. The number plate images taken from the cameras of the parking lot system have a constant size and brightness. The accuracy of the license plate recognition system installed at the entrance to the parking lot is more than 98%. However, the recognition accuracy of license plate number may drop by about 50% to 70% in video stream taken on roads, where the size and the brightness of number plate images in video stream are not constant. They are very different to each other due to the camera angle, the distance to the vehicles and the light. This situation makes some objects to be recognized as license plates compared to the constrained situation. A major cause of this low recognition accuracy is the inability to filter out noise like an outlier that has a similar pattern to license plates. More deeper and wider CNN technique can be useful to increase the accuracy in these situation. But, the IoT camera generally has a restricted resource and is unable to use too large deep CNN, which requires a lot of computation power and memory. This paper suppresses outliers, non-license plate objects and increases the recognition accuracy through time series analysis. We use the time series analysis as the statistical methods to solve this problem of the outlier occurrence. Experimental results show that the recognition accuracy in video stream taken on ordinary roads is 77% in the case of the absence of outlier filters, but 85% in the case of including the filters proposed in the paper.

Wearable haptic devices that convey skin stretch have been used in a broad range of applications, from prosthesis proprioception to language transmission. Despite their prevalence, rigorous evaluation of the perception of skin stretch cues is still ongoing. Prior studies indicate that skin stretch cue presentation velocity may impact cue perception, but we lack quantitative data regarding the impact of skin stretch velocity on cue perceptibility. It is important to understand the impact of presentation velocity to ensure the haptic cues are delivered in the most salient manner. In this paper, the Method of Constant Stimuli and Likert surveys were used to capture the just noticeable difference (JND) and participant impressions for two rotational velocities of the Rice Haptic Rocker. The velocities tested did not affect the JND; however, participants reported the faster speed was easier to discern. This study suggests skin stretch devices can be expected to maintain their perceptual performance at varying actuation speeds, meeting the requirements of a variety of applications.

Real-time object detection models based on deep learning are being studied. However, when using deep learning, the user must directly select one of the various object detection models, and the result of object detection may vary depending on the selected object detection model. Therefore, in this paper, we propose an efficient deep learning platform for object detection technology. The proposed platform estimates learning results based on benchmark results and recommends proper object detection model based on deep learning to minimizes user intervention.

In this chapter we develop a prototype framework for the development of economic models for African countries. The prototype model is an equational framework intended to describe the structural characteristics and the behavior of major macro-economic agents and variables as simultaneously determined functional relationships among key economic variables, including policy-determined values. An economy is assumed to be small and open in the sense that activities and prices in the country are affected by conditions in world markets where its own influence is minimal. Furthermore, the country is assumed to be a price-taker for its exports of primary commodities. Although the prototype model given below includes specific institutional details of an economy, additional country-specific information is incorporated during the process of model development. The prototype model is used as the basis for the construction of the empirical models for selected African economies.

The Precision Agriculture (PA) plays a crucial part in the agricultural industry about improving the decision-making process. It aims to optimally allocate the resources to maintain the sustainable productivity of farmland and reduce the use of chemical compounds. [17] However, the on-site inspection of vegetations often falls to researchers' trained eye and experience, when it deals with the identification of the non-crop vegetations. Deep Convolution Neural Network (CNN) can be deployed to mitigate the cost of manual classification. Although CNN outperforms the other traditional classifiers, such as Support Vector Machine, it is still in question whether CNN can be deployable in an industrial environment. In this paper, we conducted a study on the feasibility of CNN for Vegetation Mapping on lawn inspection for weeds. We would like to study the possibility of expanding the concept to the on-site, near realtime, crop site inspections, by evaluating the generated results.

As the Internet of Things (IoT) technology evolves, it is no longer appropriate to process real-time, large-volume data generated by numerous IoT devices in a cloud computing environment. To solve this problem, fog computing has been proposed which minimizes response time and makes real - time processing suitable. However, there is still a lack of research on techniques for efficiently managing various services and monitoring IoT devices in real time. Therefore, this paper proposes an efficient framework to monitor IoT devices in fog computing environment. The proposed monitoring framework is based on the oneM2M standard and consists of three detailed frameworks: device manager, monitoring manager, and data manager.

Skin stretch haptic devices are well-suited for transmitting information through touch, a promising avenue in prosthetic research, addressing the lack of feedback in myoelectric designs. Rocker-based skin stretch devices have been proposed for sensory substitution and navigational feedback, but the designs vary in their geometry. Other works create torsional stretch, and utilize nonlinear mappings to enhance perception. This work investigates parameters of rocker geometry and mapping functions, and how they impact user perception. We hypothesize that perceptual changes are dependent on the choice of stretch increment sizes over the range of motion. The rocker geometry is varied with an offset between the rotational and geometric axes, and three rocker designs are evaluated during a targeting task implemented with a nonlinear or linear mapping. The rockers with no offset and a positive offset (wide) perform better than the negative offset (narrow) case, though the mapping method does not affect target accuracy.

Remote Sensing has been contributing to increase in agricultural production with Site-Specific Crop Management based on development of the Geospatial technologies. In particular, identifying the crop disease affected regions becomes one of the major issues. One of methodology to solve this problem is Machine Learning technology. In this paper, we implemented an effective White-Mold disease (Sclerotinia sclerotiorum) identification with Support Vector Machine (SVM) and evaluated the accuracy using Spatial Autocorrelation (SA). Our implementation steps included relevant imagery extraction using the Spectral Vegetation Indices (SVI), SVM for White-Mold region identification using SVIs, and discernment of crop disease clustering pattern using SA. The obtained accuracy of SVM is up to 70% that showed promising performance of White-Mold identification from the given field. SA could identify a clustering pattern from given dataset, but the p-value analysis indicates that it is not statistically significant to generalize the clustering pattern onto whole White-Mold region.

This paper proposes a Decision Management System to identify the white mold regions from the soybean fields using Autologistic Statistical Model (ASM) and Remote Sensing (RS) data analysis with commercially available Big Data sets as input data. In order to develop an identification model, numerous types of data need to be considered. In this study, the data that was used is satellite image pixel values, and data gathered from the field such as precipitation, yield, elevation, humidity, wind speed, wind direction and geospatial locations. The model evaluated the outcome using this information as input parameters and provided an overall estimation of the white mold region in the soybean fields. Based on the evaluation of the result, the accuracy rate of the proposed methods 84% which is a promising result due to the fact that each pixel of the satellite image is 30 by 30 meters.

Virtual machines are used in various fields to increase the resource utilization. In a virtual machine monitor, I/O requests are the most frequent tasks for virtual machines. However, a virtual machine monitor cannot guarantee I/O responsiveness when multiple tasks are being performed on virtual machines. We propose a scheduling scheme that provides fast I/O responsiveness for virtual machines. The proposed scheme provides high resource utilization by dynamically allocating virtual resources to virtual machines according to the characteristics of their tasks. In experiments with virtual machines containing varying task characteristics, it is shown that the proposed scheme provides improved I/O responsiveness when compared to existing virtual machine schedulers.

In this paper, we present an on-line real-time physics-based approach to motion control with contact repositioning based on a low-dimensional dynamics model using example motion data. Our approach first generates a reference motion in run time according to an on-line user request by transforming an example motion extracted from a motion library. Guided by the reference motion, it repeatedly generates an optimal control policy for a small time window one at a time for a sequence of partially overlapping windows, each covering a couple of footsteps of the reference motion, which supports an on-line performance. On top of this, our system dynamics and problem formulation allow to derive closed-form derivative functions by exploiting the low-dimensional dynamics model together with example motion data. These derivative functions and their sparse structures facilitate a real-time performance. Our approach also allows contact foot repositioning so as to robustly respond to an external perturbation or an environmental change as well as to perform locomotion tasks such as stepping on stones effectively.

Image classification and extracting the characteristics of a tumor are the powerful tools in medical science. In case of breast cancer medical treatment, the breast cancer classification methods can be used to classify input images as normal and abnormal classes for better diagnoses and earlier detection with breast tumors. However, classification process can be challenging because of the existence of noise in the images, and complicated structures of the image. Manual classification of the images is time-consuming, and need to be done only by medical experts. Hence using an automated medical image classification tool is useful and necessary. In addition, having a better training data set directly affect the quality of classification process. In this paper, a method is proposed based on supervised learning and automatic thresholding for both generating better training data set, and more accurate classification of the mammogram images into Normal/Abnormal classes. The procedure consists of preprocessing, removing noise, elimination of unwanted objects, features extraction, and classification. A Support Vector Machine (SVM) is used as the supervised model in two phases which are testing and training. Intensity value, auto-correlation matrix value of detected corners, and, energy, are three extracted features used to train the SVM. Experimental results show this method classify images with more accuracy and less execution time compared to existing method.

This paper proposes a sophisticated classification process to segment the leaves of carrots from weeds. In the early stages of the plants' development, the color of both the plants and the weeds are similar, making it difficult to differentiate between the two. The process becomes even harder if the weeds and plants overlap. The proposed system addresses this problem by creating a sophisticated mean for weed identification. The major components of this system are composed of three processes: image segmentation, feature extraction and decision-making. In the image segmentation process, the input images are processed into lower units where the relevant features are extracted. In the decision-making process, the system makes use of the Support Vector Machine to analyze and segregate the weeds from the plants. Afterward, the findings are used to dictate which plants receive herbicides and which do not. The main priority for the image segmentation process is on the overlapping images where weeds need to be isolated from plants so that they can be used for cultivation purpose. The evaluation of the approach is done using an open dataset of images consisting of carrot plants. The system is able to achieve 88.99% accuracy for weed classification using this dataset. This methodology will help to reduce the use of herbicides while improving the performance and costs of precision agriculture.

Image segmentation is a challenging task in image processing. The purpose is to divide up pixels into different partitions in which members of each partition have similar characteristics and a unique label. Image segmentation and extracting the characteristics of a tumor are powerful tools that can be used in medical science. In the case of breast cancer medical treatment, segmentation methods can be used to extract and segment the tumor for better diagnoses and earlier detection of breast tumors. However, extracting and segmentation of the tumors or region of interest (RIO) can be challenging. This is due to the existence of noise in the images, along with the complicated structures of the image. Manual classification of the images is time-consuming, and needs to be done only by medical experts. Hence, using an automated medical image segmentation tool will be useful and necessary. In this paper, a method is proposed based on the well-known watershed technique and automated thresholding for single and multi-tumor segmentation in medical images. The procedure consists of pre-processing, removing noise, elimination of unwanted objects, generating and segmentation. Segmentation involves automatic thresholding, gradient magnitude, finding regional minimums, and recognition. Experimental results show that this method performs well in segmentation with efficient execution time and can be used for medical diagnostics of breast cancer.

Similarity measurements in images have always been a challenging task in the field of pattern recognition techniques. Shape based features is a widely adopted method in Content Based Image Retrieval (CBIR) for similarity measurement. In this paper we proposed an enhanced version of Mesh Distance Fourier Descriptor (MDFD) previously developed in our lab for the similarity measurement. Two extra levels of filters have been added to the output of MDFD so that the final output is more refined and the most similar image to the query image is selected from the database. The first level filter includes processing of images retrieved based on ratio of area of the image to the area of minimum bounding rectangle enclosing that image. The second level filtering includes calculation of average of absolute difference of global features like eccentricity, convexity and solidity of the query image and retrieved image. Adding these two extra filtering levels, the matching ratio has been increased from 84% to 88% which shows adding filters enhances the results of MDFD. In this paper we have used binary images extracted from region of interest (ROI) of mammogram which are classified into single objects using known classification methods such as K-means and SVM algorithms.

Automated vehicle detection and classification is one of many highly researched areas in Computer Science. This paper proposes and presents a hybridized method for classifying objects based on spatial orientation. Specifically, the proposed classification system explores the Histogram of Oriented Gradients feature extractor conjoined with a clustering algorithm to classify vehicle images in an unsupervised manner. HOG is a well-suited feature extractor for dense images rich with contours and edges. The pairing provides an efficacious orientation classifier for vehicle images. Training and sample data exceeded 1.8 million images and was provided by Carsforsale.com, Inc.'s vast catalogue of historical vehicle images.

In this paper, we present a shape based image retrieval method based on chain code representing irregularity of an object. A distinctive chain code is introduced as a main extracted feature of the object. All objects used in this paper are binary object images extracted by well-known classification algorithm, Support Vector Machine (SVM). From these classified binary images, we propose a modified shape based image retrieval method with the unique chain code interpreting irregularity of object. Proposed method is experimented along with known shape based image retrieval method using characteristic point features. The experimental result shows that proposed method exceed matching rate that of conventional contour to centroid triangulation (CTCT) method showing proposed method has higher matching rate.

In this paper, we propose a new image retrieval method based on Sectored Contour to Centroid Triangulation (SCTCT) using distinctive shape feature, named Arc Difference Rate (ADR). We utilized Support Vector Machine (SVM) method as an extraction tool to extract suspicious tumor area as binary object image from the breast MRI. Therefore extracted 100 binary object images are used as test cases in the experimental study. The results from proposed method show the improvement in finding correct matches compare to the traditional SCTCT.

Multimedia applications based on cloud services for mobile devices have recently gained considerable popularity. However, the increasing density of devices leads to a high level of interference, which reduces the performance of wireless communication between devices and cloud. In this paper, we propose a new approach which allows the network to adaptively find a transmission opportunity scheduling strategy by choosing the most valuable transmission request opportunity. In this approach, the transmission request selection strategy is optimized by considering multimedia distortion reduction, hidden node problem, transmission interference, and signal coverage. Simulation results show that the proposed request selecting strategy significantly improves the systems overall data transmission quality by exploring the tradeoff between communication node pair and their neighbor nodes.

Recently, it becomes increasingly important to secure user private data in mobile devices. To protect user private data, one possible approach is to implement a secure file storage in the mobile devices based on mandatory access control (MAC), but the device manufacturers seldom implement it because of high pressure of time-to-market, frequent version upgrade of the operating system, and no unanimous agreement in the MAC standard software. In this paper, we propose an implementation study of a secure file storage, called a ghost drive, which can facilitate the implementation of MAC in the mobile devices by unburdening the manufacturers from aggressive instrumentation of the whole operating system. Since our implementation is in form of a loadable kernel module, separated from the main kernel, it can be deployed even to commercial mobile devices already in use by installing it over the air. Our experiments show that the performance of our secure storage implementation is not worse than the original unmodified implementation.

About 1 in 8 women in the United States is expected to develop breast cancer over the course of herentire lifetime but a few medical imaging techniques have been applied for breast cancer screening. In addition, the feature extraction and comparison in medical images for breast cancer detection haverarely been reported in literature. We propose a new framework toextract agglomerated features in medical imagesand comparethem by relating original characteristic patterns thereof. Our method concentrates on three key aspects and they are: a comparison between intensity distributions of pixels collected by the hexagonal mask, detecting minimum gradient points in a radial intensity series, and generatinga characteristic pattern of the feature. The main contribution of ourproposed approach is improving a method of identifying features which is lesssensitive to noise in medical images for breast cancerdetectionand presenting an original design of relating features which is consistent to the orientation and size of the feature. Experimental results demonstrate that our proposed approach is more tolerant of image noise than prior research and generates an invariant characteristic pattern of various orientations and sizes.

The segmentation of breast Magnetic Resonance Imaging (MRI) has been a long term challenge due to the fuzzy boundaries among objects, small spots, and irregular object shapes in breast MRI. Even though intensity-based clustering algorithms such as K-means clustering and Fuzzy C-means clustering have been used widely for basic image segmentation, they resulted in complicated patterns for computer aided breast MRI diagnosis.In this paper, we propose a new segmentation algorithm to improve the clustering results from K-means clustering algorithm with breast MRI. The major contribution of the proposed algorithm is that it simplifies breast MRI for the computer aided object analysis without loss of original MRI information. The proposed algorithm follows K-means clustering algorithm and explores neighbors and boundary information to redistribute unexpectedly clustered pixels and merge over-segmented objects from K-means clustering algorithm. We will discuss the results from the proposed algorithm and compare them with the result of K-means clustering algorithm.

Due to the benefits of its reusability and productivity, the component-based approach has become the primary technology in service robot software frameworks, such as MRDS (Microsoft Robotics Developer Studio), RTC (Robot Technology Component), ROS (Robot Operating System) and OPRoS (Open Platform for Robotic Services). However, all the existing frameworks are very limited in fault tolerance support, even though the fault tolerance function is crucial for the commercial success of service robots. In this paper, we present a rule-based fault tolerant framework together with widely-used, representative fault tolerance measures. With our observation that most faults in components and applications in service robot systems have common patterns, we equip the framework with the required fault tolerant functions. The system integrators construct fault tolerance applications from non-fault-aware components by declaring fault handling rules in configuration descriptors or/and adding simple helper components, considering the constraints of the components and the operating environment. Much more consistency in system reliability can be obtained with less effort of system developer. Various fault scenarios with a test robot system on the proposed OPRoS fault tolerant framework demonstrate the benefits and effectiveness of the proposed approach.

This paper proposes and measures the performance of a hybrid prefetch scheme for the gateway cache in the vehicular telematics network. Aiming at improving response time and instant reply ratio, the proposed scheme combines the advantage of LRU and FAR to best exploit temporal and spatial locality. For the advanced path recommendation service, links are categorized into two groups, namely, one referenced during the given interval, and the other not referenced. Each group orders the links according to the distance from the gateway, finalizing the prefetch order. Simulation results, based on the real-life trajectory data obtained from a real-time tracking system, show that the hybrid prefetch can improve instant reply ratio by up to 5.3% and response time by up to 12.5%, compared with FAR in spite of low link-level hit ratio.

In this paper we introduce the application of partitioning-based SOR (PSOR) algorithm to the minimum energy multicast tree problem in wireless ad hoc networks. PSOR was proposed for the minimum energy broadcast tree problem in our previous work and it can be applied to the multicast tree problem with small modification. PSOR for multicast maintains the same theoretical bound for the approximation ratio and computational running time as PSOR for broadcast. The computational results show improvement in the solution quality (in terms of the sum of the transmission powers). PSOR outperforms the other algorithms and it shows better results compared to ISOR (Iterated version of SOR) in terms of the ratio of solution quality improvement over running time increase. Recent careful study revealed that there was a bug in our implementation of SOR which caused worse performance than it should have. In this paper we used the results of the corrected version of SOR for the comparisons.

In this paper, we present an enhancement to our previously proposed algorithm, SOR, for the minimum energy broadcasting problem in wireless sensor ad hoc networks. We implement the enhanced algorithm, PSOR (Partitioning- based SOR), and compare its solution quality with other algorithms in the literature: BIP, OMEGa, SOR, and EWMA. The enhancement comes from the diversification of solution space by allowing more shrinking than the original algorithm, SOR. We can achieve the diverse solution search without hurting the original algorithm's theoretical performance bound. The experimental results confirm that the enhancement further improves the solution quality of the original algorithm, SOR.

Welcome to the Software Engineering Track of ACM SAC 2008 -- The 23nd Annual ACM Symposium on Applied Computing. This year we have a special focus on "Emerging Techniques for Dependable Software Development". The objective is to provide researchers and practitioners with an embedded forum for presenting and discussing their ideas and experiences with technologies, theories, and tools used for producing highly dependable software more efficiently asnd effectively. The Software Engineering Track is big enough to represent major topics in software engineering, but also small enough to provide an in-depth representation of theory and practice in these areas. It provides an opportunity for academic researchers and industry participants to share their ideas and practices. It not only allows the software engineering academic community to understand the areas that are vital to the software industry, but it also gives software engineering practitioners an opportunity to express their needs.

In most file systems, if a file is deleted, only the metadata of the file is deleted or modified and the file's data is still stored on the physical media. Some users require that deleted files no longer be accessible. This requirement is more important in embedded systems that employ flash memory as a storage medium. In this paper, we have designed a NAND flash file system that has a secure deletion functionality. We modified YAFFS to support secure deletion. Our method uses encryption to delete files and forces all keys of a specific file to be stored in the same block. Therefore, only one erase operation is required to securely delete a file. The proposed method securely deletes not only keys but also all of the metadata of that file. Our simulation results show that the number of block erases due to file creation and file modification is very low and the amortized number of block erases is lower than the simple encryption method. Even though we applied our method only to the YAFFS, our method can be easily applied to other NAND flash file systems.

Loadable kernel modules supported by Linux provides lots of benefits such as a small-sized kernel, on-demand loading, and easy software upgrading. However, since modules are executed in a privileged mode, trivial misuses in a module may cause critical system halts or deadlock situations. This paper presents a kernel resource protector which prevents kernel from faults generated by modules. The protector models the system in two objects: module object and resource object. By observing the interrelations between the two objects, the protector can detect misuses of modules and take actions to resolve the erroneous situations. Implementation study has shown that the protector can find out memory leaks wasted by modules and can reclaim leaks without degrading system performance. The proposed protector makes Linux more robust, which is required indispensably in the system equipped with NVRAM (Non Volatile RAM) such as FRAM and PRAM.

Incremental checkpointing, which is intended to minimize checkpointing overhead, saves only the modified pages of a process. However, the cumulative size of incremental checkpoints increases at a steady rate over time because many updated values may be saved for the same page. In this paper, we present a comprehensive overview of Pickpt, which is a page-level incremental checkpointing facility. Pickpt provides space-efficient techniques for minimizing the use of disk space. For our experiments, the results show that the use of disk space of Pickpt was significantly reduced compared with existing incremental checkpointing.

We compared the performance of Bayes Theory, Fuzzy Set Theory, Heuristic method and Dempster-Shafer Theory in the identification of aircrafts by using the information from different sensors. From the results of the simulation, the Fuzzy method produces the best result. The final identification could be improved if specific features of each type of aircraft are available. To get more accurate object identification, the results from each method can be combined with the best assignment of values for each method.

In this paper, we develop a security design metric that can be used at system design time to build more secure systems. This metric is based on the system-wide approach and adopt a reliability model and scenario testing technique to produce a feedback cycle.

In general, the interactions between proteins are fundamental to a broad area of biological functions. In this paper, we try to predict protein-protein interactions in parallel on a 12-node PC-cluster using domains of a protein. For this, we use a hydrophobicity among protein's amino acid's physicochemical feature and a support vector machine (SVM) among machine learning techniques. According to the experiments, we get approximately 60% average accuracy with 5 trials and we obtained an average speed-up of 5.11 with a 12-node cluster using a proximal SVM.

In this paper, we introduce the application of partitioning-based SOR (PSOR) algorithm to the minimum energy multicast tree problem in wireless ad hoc networks. PSOR for broadcast was proposed in our previous work and it can be applied to the multicast tree problem with small modification. PSOR for multicast maintains the same theoretical bound for the approximation ratio and computational running time as PSOR for broadcast. The computational results show improvement in the solution quality (in terms of the sum of the transmission powers). PSOR outperforms the other algorithms and it shows better results compared to ISOR (iterated version of SOR). In this paper, we used the results of the corrected version of SOR for the comparisons. In addition, we present an improved version of PSOR based on simplified partition recovery method. The simulation results show significant reduction of computation time by the improved version.

A technique is described for automatic validation of communication protocols that are given in the form of regular protocol diagrams. It is based on the use of the link state introduced by West [2], but it has the advantage of not requiring that the state transition diagrams of the two communicating processes be available. The technique is applied to the validation of the CCITT X.21 protocol to illustrate the ease with which it can be used to detect reception errors, deadlock errors and overflow errors.

This paper presents some formal generative specifications for describing the graph languages consisting of the sets of directed graphs with labeled arcs. Some mathematical results are introduced. In addition, we consider applications to specification technique for data structures and complexity measure of computer software. Finally, directions for future research are indicated.

Many software systems adopt isolation mechanisms of modern processors as software security building blocks. Reasoning about these building blocks means reasoning about elaborate assembly code, which can be very complex due to the loose structure of the code. A way to overcome this complexity is giving the code a more structured semantics. This paper presents one such semantics, namely a fully abstract trace semantics, for an assembly language enhanced with protection mechanisms of modern processors. The trace semantics represents the behaviour of protected assembly code with simple abstractions, unburdened by low-level details, at the maximum degree of precision. Additionally, it captures the capabilities of attackers to protected software and simplifies providing a secure compiler targeting that language. ispartof: pages:1562-1569 ispartof: Proceedings of the 29th Annual ACM Symposium on Applied Computing pages:1562-1569 ispartof: SAC location:Gyeongju date:24 Mar - 28 Mar 2014 status: published

We introduce an extension of linear constraints, called linear- range constraints, which allows for (meta-)reasoning about the approximation width of variables. Semantics for linear- range constraints is provided in terms of parameterized linear systems. We devise procedures for checking satisfiability and for entailing the maximal width of a variable. An extension of the constraint logic programming language CLP(R) is proposed by admitting linear-range constraints.

International audience; Web office suites such as Google Docs offer unparalleled collaboration experiences in terms of low software requirements, ease of use, data ubiquity, and availability. When the data holder (Google, Microsoft, etc.) is not perceived as trusted though, those benefits are considered at stake with important privacy requirements. Content cloaking is a lightweight, cryptographic, client-side solution to protect content from data holders while using web office suites and other "Web 2.0", AJAX-based, collaborative applications.

International audience; As large-scale parallel platforms are deployed to comply with the increasing performance requirements of scientific applications, a new concern is getting the attention of the HPC community: the power consumption. In this paper, we aim at evaluating the viability of using low-power architectures as file systems servers in HPC environments, since processing power is of less importance for these servers. We present a performance and energy efficiency study of such low-power servers when compared to conventional architectures. Our results indicate that the low-power alternativecould be a viable choice to save energy by up to 85% while not compromising on performance, specially for read-intensivework-loads. We show the low-power server provides 7 times more energy efficiency to the system while running a real application from the seismic wave propagation field.

International audience; Active objects are programming constructs that abstract dis- tribution and help to handle concurrency. In this paper, we extend the multiactive object programming model to offer a priority specification mechanism. This mechanism allows programmers to have control on the scheduling of requests. The priority representation is based on a dependency graph which makes it very convenient to use. This article shows how to use this mechanism from the programmer side, and exposes the main properties of the dependency graph. The software architecture of our implementation is also presented, as it can be applied to various scheduling systems. Finally, we validate our approach through a microbenchmark that shows that the overhead of our priority representation is rather low. On the whole, we provide a general pattern to introduce a prioritized scheduling in active objects or in any other con- current systems. The resulting framework is shown to be fine-grained, user-friendly, and efficient.

International audience; This paper presents a process calculus specifically designed to model systems based on the Internet of Things paradigm. We define a formal syntax and semantics for the calculus, and show how it can be used to reason about relevant examples. We also define two notions of bisimilarity, one capturing the behavior seen by the end user of the system, and one allowing compositional reasoning.

Session: Volume II: Software development and system software & security: software verification and testing track; International audience; Various combinations of static and dynamic analysis techniques were recently shown to be beneficial for software verification. A frequent obstacle to combining different tools in a completely automatic way is the lack of a common specification language. Our work proposes to translate a Pre-Post based specification into executable C code. This paper presents e-acsl, subset of the acsl specification language for C programs, and its automatic translator into C implemented as a Frama-C plug-in. The resulting C code is executable and can be used by a dynamic analysis tool. We illustrate how the PathCrawler test generation tool automatically treats such pre- and postconditions specified as C functions.