Many real world problems have iterative behaviors, so planners should be able to tackle such problems and generate iterative plans. Iterative actions have been considered by a few of planners, and most of the efforts have been based on linear planning. In this paper, we show how SAT-Plan, which is one of the fastest existing planners, can be extended to generate recursive plans. Generated recursive plans are similar to those that are generated by humans for iterative problems. The main idea in this work is based on the mathematical induction, and has been implemented. Results obtained from testing the extended SAT-Plan on several examples from the blocks-world domain, confirms the success of the proposed idea.

Conventional software validation methods for real-time and reactive programs are not fully reliable. Considering complexity of formal verification and incompleteness of testing approaches, runtime verification approach is used. In this paper, Safety properties (after extracting from program requirement specification) are represented in Real-Time Logic (RTL) and fault tolerant module reaction are added to the program. Program monitoring module (which is realized by Functional, Timing and Deadline Aspects) is weaved to the program code. The enriched program executes in the environment. The monitoring module verifies program behavior by considering the specified safety properties and when any violence of a property is detected, then the specified reaction (Fault Tolerance) activates and navigates the program to a safe state.

Variability management is a major concern in successful exploitation of variabilities and commonalities of software product families which also affects different aspects of development activities. To use software product line approach in information systems context, it is necessary to bring in variability in different phases of the life cycle, including business modeling which is recognized as a key part of developing enterprise information systems. This paper presents a method for variability management at business process level which covers the activities of eliciting variability, representing variability in business process models, and also managing dependencies among different variabilities. The method supports traceability by mapping variabilities at business process level to assets of lower levels of abstraction, including feature and use case models. To model business processes, UML activity diagrams are exploited.

Real-time video transmission is considered as an important way for information broadcasting. One noticeable example in this area is E-leaning applications, which is based on real-time video processing and transmission. On the other hand, the process of registration is a fundamental component in automatic image and video processing. In our previous work, we implemented a new method for cut detection technique based on dominant lines and angles. This paper introduces a new video registration technique that uses dominant angles extracted from edge information of the video frames in one shot. To the best of our knowledge, it is the first works done for registration in E-learning applications. This method is compatible with our application’s requirements and has a low complexity and high speed. We compare our method against three established techniques and present our results using different video sequences.

This paper presents a computationally effective trajectory generation algorithm for omni-directional mobile robots. In this algorithm we use the Voronoi diagram to find a sketchy path that keeps away from obstacles and then we smooth this path with a novel use of Bezier curves. This algorithm defines velocity magnitude of a robot along the curved path to satisfy optimality conditions and dynamic constrains. The algorithm has been implemented on real robots, and we present experimental results in different environments.

The security of networked computers plays a strategic role in modern computer systems. This task is so complicated because the determination of normal and abnormal behaviors in computer networks is hard, as the boundaries cannot be well defined. This paper proposes a parallel genetic local search algorithm (PAGELS) to generate fuzzy rules capable of detecting intrusive behaviors in computer networks. The system uses the Michigan’s approach, where each individual represents a fuzzy rule which has the form “if condition then prediction”. In the presented algorithm the global population is divided into some subpopulations, each assigned to a distinct processor. Each subpopulation consists of the same class fuzzy rules. These rules evolve independently in the proposed parallel manner. Experimental results show that the presented algorithm produces fuzzy rules which can be used to construct a reliable intrusion detection system.

Error Back Propagation, a class of neural networks, is proposed to solve the inverse kinematics problem in robotic manipulator. In this approach a network has been trained to learn a desired set of joint angles positions from a given set of end effectors positions. This paper demonstrates some methods of Back Propagation neural network which can be used to solve inverse kinematics. Next the performance of these methods has been compared for inverse kinematics problems. The used Error Back Propagation techniques are the Standard, Momentum and Delta Bar- Delta.

The flexible DFT strategy helps designers control the eventual cost of test during the chip design phase. To reach a uniform test strategy for CPU data path, we use S-graph information. But register files and internal memory structures cannot be easily represented by S-graphs. In most processors investigated, one can find some sort of internal memory like general-purpose registers, stacks or queues. The control hardware and addressing schemes of such structures make it difficult to test them. We design a wrapper around these structures to isolate them from data path and incorporate them to S-graphs applications. These compatible S-graphs provide a uniform BIST strategy for the whole data path. The wrapper design can test itself concurrently with other modules so it can reduce the test application time. We apply our method on SAYEH CPU as a vehicle.

This paper is about the vision and odometry errors of a sweeper mobile robot. The robot takes advantage of color recognition in its vision in order to detect different types of objects having different colors. An effective artificial neural network which is extremely easy to implement and is surprisingly quick in practice for mobile robots is introduced. Consequently, the connectionist is applied on the robot for object detection and the gained results are compared with other methods. Also experimental tests plus statistical analyses are carried out in order to measure the robot localization errors.

Several methodologies with their own characteristics have been proposed in the area of agent-oriented software engineering. Consequently, deciding which methodology to select in a specific case is an important issue and it can lead to decrease software development cost and effort. Thus, importance of evaluation of methodologies will be highlighted in choosing the appropriate methodology in the development process of an application. It can also help in developing new methodologies and improving existing ones. In this paper, we are going to provide an evaluation framework of agent oriented methodologies. To demonstrate the usage of the suggested framework, it is applied to evaluate two methodologies (MESSAGE and Prometheus) using a proper example. Results show that, using our method, methodologies can be truly compared and evaluated.

Gait Recognition refers to automatic identification of an individual based on his/her style of walking; it's a new biometrics recognition technology. This paper describes a new approach to gait recognition based on kmean clustering algorithm. Body silhouette is extracted by a simple background subtraction, and the clustering is performed to partition image sequence into clusters, so the vectors of feature can be extracted. The recognition is achieved by dynamic time warping technique. We evaluate the proposed gait recognition method on the Gait Challenge database of the University of South Florida (USF), and the experimental results demonstrate that our approach has a good recognition performance.

One of the most challenging problems of pervasive computing is protecting the privacy of individuals' data by respecting the principles of awareness and consent with minimum distraction. A suitable approach should consider the intention of all authorized entities determining data disclosure and usage regulations. Therefore, as an extension to current privacy protection approaches which consider only one determiner for a private data item, we propose the idea of multi-determiner privacy protection. In this paper, we investigate possible determiners of private data and propose a set of meta-data which are required to cover the concept of multideterminacy including ownership-information, operators, and DELEGATION behavior. Besides, we devise and evaluate a deadlock-free and conflict-free algorithm to manage ownership and DELEGATION scenario in answering a request. We subsequently integrate the algorithm and the required meta-data in a multi-determiner privacy protection architecture which works as a mediator to protect multi-determiner private data. We demonstrate the applicability of our work through a prototype implementation and realization of a sample scenario.

Threshold setting methods are one of the most important tasks in any speaker verification system. Equal Error Rate (EER) and Minimum Error Rate (MER) criteria are usually used for performance evaluation in speaker verification systems. It is important to know that EER and MER are two a posteriori methods for decision threshold estimation and can not be used in real world applications. In real world applications decision thresholds should be determined a priori, i.e. a validation phase is needed to conduct some inter-speaker and intra-speaker verification tests and to use obtained distances or likelihoods for estimation of a priori decision thresholds. In this paper, we present a least square error based method for calculation of parameters which we need to estimate decision thresholds. For this we use means and variances of inter-speaker and intra-speaker distances and estimate necessary parameters for calculating decision thresholds. The proposed method leads to an optimum a priori estimation of decision threshold values.

Normalization methods are used to improve accuracy in speaker verification systems. In this paper, some of the most usual normalization methods including maximum, TNorm, and ZNorm normalization methods were investigated on GMM speaker modeling approach. Maximum normalization method was shown to have the best accuracy among the studied methods. Also, hybrids of normalization methods led to better speaker verification accuracy. In this paper, serial hybrid of normalization methods was also considered. It was shown that applying Maximum normalization method for two successive times leads to the best result. Serial hybrid approach increased recognition time. So we proposed twostage normalization approach that has special consideration on accuracy and recognition time of speaker verification system.

Attacks on Computer networks can be divided into four groups including denial of service (DOS), unauthorized access from remote machine (R2L), unauthorized access to super user privileges (U2R) and probing (Probe). Using information collected from a system along probing attack, an attacker can identify vulnerability of victim system and specify the kind of attack to penetrate that system. Probing attack is the preamble of other attacks, so if that attack is detected and prevented, the immunity of the system is more promoted . In this paper we use support vector machine (SVM) to detect probing attack. The results of our experiments performed on dataset provided by the DARPA intrusion detection evaluation program; show that SVM classifier outperforms many other classifiers for detection of probing attacks.

In 2000, Ekdahl and Johansson introduced the stream cipher SNOW, as a proposal for the NESSIE project, but a few Guess and Determine (GD) attacks followed and indicated certain weaknesses in the design. Then a new version of SNOW, called SNOW2.0, was developed as a modified version of SNOW1.0. Yet this stream cipher is also vulnerable against some introduced attacks. This paper gives some criteria of modifying an LFSR-based stream cipher against GD attacks. Next, using one of these criteria, we introduce a modified version of SNOW2.0 with respect to GD attacks. The results of evaluating the modified SNOW2.0 against other general attacks show that the new proposed algorithm is more resistant against other types of general attacks, i.e. TMDTO, distinguishing and algebraic attacks.

In our paper, a new approach for critical success factors of Business to Business (B2B) electronic commerce is presented. Important aspects of this approach can be classified into three main groups; environmental, internal and inter-organizational factors. The first group pertains to the factors that change the conditions equally for all companies in the business space and are out of influence of firms. Internal factors comprise individual features of companies, such as business strategies, culture, resources and etc. The Inter-organizational dimension is included the factors that are related to organization ability to set up Inter-organizational relationships electronically. Experimental results showed promising in evaluating the quality of business enterprises.

Modulo 2n +1 adders are important for several applications including residue number system implementations, digital signal processors and cryptography algorithms. In this paper we present a new number system and a novel addition algorithm for its operands. In this paper, we present two new architectures for designing modulo 2n +1 adder, based on ripple carry adder. The first architecture utilizes a more rapid architecture whereas the second applies less hardware. In the proposed method, the special treatment required for zero operands in diminished-one number system is removed. In the fastest modulo 2n +1 adders in normal binary system, we are faced with 3-operand adders. This problem is also resolved in this paper.

The assignment of natural language texts to two or more predefined categories based on their contents, is an important component in many information organization and management tasks. This paper presents an information theoretic approach for text classification problem that we call it ITTC. Here, we prove that ITTC is theoretically equivalent to Bayesian classifier. However, when classification task is performed over dynamic or noisy data, or when the training data do not represent all probable cases, ITTC outperforms Bayesian classifier. We also show that the complexity of ITTC over test set grows linearly by the size of input data .We use some news groups, to evaluate the superior performance of our approach.

In this paper, the application of “Multivariate Adaptive Regression Splines” (MARS) to the problem of segmental (phonemic) duration modelling in Farsi text-to-speech systems is presented. Segmental duration is influenced by a number of contextual factors such as segment identity, stress, position of a target segment within a syllable, word, and phrase. These factors interact with each other and a good model of segment duration should account for the problem of factor's interaction. Databases of speech data often encounters with sparse data problem. MARS is a technique to estimate general functions of high-dimensional arguments given sparse data, which automatically selects the parameters and the structure of the model based on available data and deals with the problem of interaction between factors. Besides highly accurate prediction, a MARS model also allows interpretation of its structure. Using MARS method for Farsi segmental duration modeling yields a correlation coefficient of 86.50 between observed and predicted durations for training data of and a correlation coefficient of 80.83 between observed and predicted durations for testing data. The performance of MARS model was also compared to Multi-Layer Perceptron (MLP) neural network. MLP neural netwok was trained using an error Back Propagation algorithm. Using MLP neural network for segmental duration medeling of Farsi language leads to a model with a correlation coefficient between observed and predicted durations of 84.86 for training data and 80.97 for testing data.