Exploring AdaBoosting Algorithm for Combining Software Reliability Models
Track Name:
Student Papers
Abstract:
Software reliability growth models (SRGMs) are very important for software reliability estimation and prediction and have been successfully applied in software reliability engineering. However, there is no general model which can perform well for different cases. Therefore, several combinational methods of SRGMs have been proposed to improve the reliability estimation and prediction accuracy. AdaBoosting (short for Adaptive Boosting) is a commonly used machine learning (ML) algorithm for combining several weak predictors into a single strong predictor to significantly improve the forecasting accuracy. In this paper, an AdaBoosting-based approach for obtaining a dynamic weighted linear Combinational Model (ACM) is proposed. The key idea of this approach is that we select several SRGMs as the weak predictors and use AdaBoosting algorithm to determine the weights of these models for obtaining the final linear combinational model. Finally, two case studies are presented respectively for comparing the proposed ACM with five SRGMs and a neural-network-based combination approach on several real failure data-sets. Preliminary experimental results show that this proposed ACM is fairly effective and applicable since it has: 1) a significantly better goodness-of-fit and prediction capability than the five single SRGMs; 2) a comparable at least or even a little better fitting and prediction performance than the dynamic weighted combinational model based on neural-network in most cases.
Author Names:
Haifeng Li*, Beihang University Min Zeng, Minyan Lu, BeiHang University
Author Emails :
lihaifeng AT dse.buaa.edu.cn studyzm AT 163.com lmy AT buaa.edu.cn
Automated Stress Testing of Windows Mobile GUI Applications
Track Name:
Student Papers
Abstract:
Mobile devices are now a common tool for personal, business, entertainment and communication use. The increase in usage of Smartphone devices has led to the increase in development of mobile software applications. The testing of mobile applications is made difficult as there are many different types of devices available that may be targeted by the same application. In addition to this, mobile devices are a resource constrained device, with less memory and processing power than the desktop PC. However, even with these constrained resources, performance and robustness are still an important factor to consider when developing mobile software applications. This research paper discusses the current status and proposed work of a research project involving the use of artificial intelligence, fuzz-testing and automated GUI testing techniques for Windows Mobile devices. It describes the current development status of an automated GUI stress testing tool, Torqueo. Then, it details the various types of scenarios for which this tool has been utilized and also outlines the advantages and limitations of this research study.
Author Names:
Nizam Abdallah*, Monash University Sita Ramakrishnan, Monash University
Author Emails :
nizam.abdallah AT infotech.monash.edu.au sita.ramakrishnan AT infotech.monash.edu.au
Subject Areas:
Reliability, Analysis and Measurement Validation, Verification and Testing*
Method for Reliability Estimation of COTS components based Software Systems
Track Name:
Student Papers
Abstract:
In this paper a relatively simple and implementable approach is proposed for reliability calculation of software systems which are constructed from COTS components. This approach estimates overall software system reliability based on the individual component reliabilities which are integrated to form the system. Moreover those components having longer execution time contribute more towards overall system reliability. Here we have used operational profile of a component for the calculation of component usage ratio, which uses operational profile for a very different purpose from its traditional use in software reliability engineering.
Author Names:
Vivek Goswami*, DA-IICT Yashwant Acharya, Physical Research Laboratory
Author Emails :
vivekgoswami10 AT gmail.com yacharya AT prl.res.in
An Analytical Framework of Survivability Model for VoIP
Track Name:
Student Papers
Abstract:
Nowadays Voice over Internet Protocol (VoIP) has become an evolutionary technology in telecommunications, and hence it is very important to study and enhance its survivability measures. In this paper, an analytical framework of survivability model for VoIP is proposed. The study is mainly focused on analyzing the combined effects of resource degradation and security breaches on Quality of Service (QoS) of VoIP, to enhance its overall performance. Software rejuvenation methodology is adopted as a preventive maintenance policy to prevent or postpone software failures. The VoIP system is modeled and analyzed as a stochastic process based on semi-Markov model to capture the effects of time spent at various states of the system. The model analysis indicates the feasibility of our approach. In addition, a comparison is made between the performance of our model with the existing models, and it is observed that our model provides better results.
A Study on Software Reliability Engineering Present Paradigms and its Future Considerations
Track Name:
Student Papers
Abstract:
Software reliability engineering is focused on engineering techniques for developing and maintaining software systems whose reliability can be quantitatively evaluated. In order to estimate as well as to predict the reliability of software systems, failure data need to be properly measured by various means during software development and operational phases. Although software reliability has remained an active research subject over the past 35 years, challenges and open questions still exist. The various modeling technique for Software Reliability is reaching its prosperity, but before using these technique, we must carefully select the appropriate model that can best suit our case. Measurement in software is still in its infancy. No good quantitative methods have been developed to represent Software Reliability without excessive limitations. In this paper, we review the introduction of software reliability engineering, the current trends and existing problems and specific difficulties, possible future directions and promising research subjects in software reliability engineering are also addressed.
Author Names:
Deepak Pengoria*, VIT University Saurabh Kumar,
Author Emails :
deepakpengoria2005 AT vit.ac.in saurabh4vit AT gmail.com
Subject Areas:
Empirical Studies* Reliability, Analysis and Measurement
A Rule Set to Detect Interference of Runtime Enforcement Mechanisms
Track Name:
Student Papers
Abstract:
Runtime enforcement aims at verifying the active execution trace of executing software against formally specified properties of the software, and enforcing the properties in case that they are violated in the active execution trace. Enforcement mechanism of individual properties may interfere with each other, causing the overall behavior of the executing software to be erroneous. As the number and the complexity of the properties to be enforced increase, manual detection of the inferences becomes an error-prone and effort-consuming task. Hence, we aim at providing a framework for automatic detection of interferences. As the initial steps to create such a framework, in this paper we first provide formal definitions of an enforcement mechanism and enforcement operators. Second, we define a rule set to detect the interference among properties.
Author Names:
Somayeh Malakuti*, University of Twente Christoph Bockisch, University of Twente Mehmet Aksit, University of Twente
Author Emails :
malakutis AT ewi.utwente.nl c.m.bockisch AT ewi.utwente.nl m.aksit AT ewi.utwente.nl
Subject Areas:
Validation, Verification and Testing* Reliability, Analysis and Measurement
A comparison of three alternative means for safety critical control
Track Name:
Student Papers
Abstract:
This article elaborates upon the strengths and weaknesses tied to three alternative controllers. The context is safety critical control applications. The three kinds of controllers are the human controller, the conventional programmable controller, and the adaptive programmable controller. The safety aspect is particularly emphasised due to the strict requirements put on critical systems for documentable evidence that a certain level of safety is achieved.
Combining Multiple Learners Induced on Multiple Datasets for Software Effort Prediction
Track Name:
Student Papers
Abstract:
Background: First approaches in software effort prediction depended on regression based models, whereas later models investigated more sophisticated methods like machine learning algorithms. Discussion Points: Single methods or models can discover only a certain part of the high dimensional space of software effort data and a common practice to increase accuracy values is to combine multiple learners. However, merely comparing models over a single dataset on the basis of precision values is not a healthy practice, since each dataset may favor a certain method. Therefore, a solid statistical test is required for comparison. Method: In this study, we adapt a previous study conducted by Khosgoftaar et. al. [1] in the field of software quality analysis to the field of software effort estimation and evaluate our results on the basis of statistical significance tests. Conclusions: Khosgoftaar et. al.’s work[1] was the first of its kind in software quality and we adapted their novel work to software effort prediction. We exploited 14 methods over 3 different software effort prediction datasets under 4 different scenarios and observed similar results to Khosgoftaar et. al., that is multiple learners induced on single dataset does not produce significantly better results.
Author Names:
Ekrem Kocaguneli*, Bogazici University Ayşe Bener, Boğaziçi University Yigit Kultur, Bogazici University
Author Emails :
ekrem.kocaguneli AT boun.edu.tr bener AT boun.edu.tr yigit.kultur AT boun.edu.tr
Selection of Fuzzy Logic Mechanism for Qualitative Software Reliability Prediction
Track Name:
Student Papers
Abstract:
The Proof of Concept (PoC) is usually considered a milestone on the way to a fully functioning prototype. A company presented with a project or proposal will often undertake internal research initially, to prove that the core ideas are workable and feasible, before going further. In general, any predictions about real time implementation are solely based on the prototype studies. A unique take on strengthening the role of the protoype itself, without actually realizing it, would be to arrive at predictions using historical information from similar PoCs or the permeating experience of those involved in projects of comparable nature. Abundance of soft computing techniques should make this crucial bypassing feasible. The purpose of the this work is to demonstrate the same. Validation of this approach could be obtained by comparing the results with the ones obtained on realized prototypes at module level. In a work of the first of its kind involving studies at the PoC level, qualititave predictions for the metric ‘number of defects’ are obtained using a generic Fuzzy Logic based modeling. A sound mathematical base for the calculation of slopes of various Fuzzy membership functions employed is explained in detail for the case studies considered. This framework is applicable to any of the process oriented developmental systems like RUP.
Ad hoc routing protocols has been designed to provide efficiency of routing services without considering security in mind. Therefore, they are vulnerable to a variety of attacks which affect the reliability of data transmission. The question is no longer to find an optimal route to a given destination but the safe route free from malicious attackers. Several secure ad hoc routing protocols, proposed in the literature, are based on public key cryptography that consumes much more resources and decreases consequently network performances. In this paper, we propose a secure routing scheme for AODV, called (SRS_AODV), which combines the hash chains and digital signatures to provide the same level of security as ARAN [8] while reducing the costs of hop-by-hop signature generation and verification. We use the ns2 simulator to analyze our protocol in different scenarios and to demonstrate its effectiveness to detect and isolate attackers.
Author Names:
Rym Guibadj*, ESI Sara Mehar, ESI Thouraya Bouabaan - Tebibel, ESI
Author Emails :
g_nesrym AT hotmail.com s.mehar123 AT gmail.com t_tebibel AT esi.dz
Subject Areas:
Security*
Paper ID:
270
Title:
Researches on the Multi-ontology based Avionics Electronics Systems Software Requirements Elicitation Method
Track Name:
Student Papers
Abstract:
Nowadays, there are many researches on the ontology-based requirements elicitation approach to obtain high quality requirements. However, there are still some disadvantages. One of the disadvantages is lacking of researches on aviation electronics systems ontology. Another is the existing ontology-based requirements elicitation approaches do not involve dynamic ontology knowledge and requirements error knowledge. This paper adopts ontology to construct the requirements knowledge multi-ontology framework of aviation electronics system. Moreover it introduces the software requirements error pattern which is merged into requirements knowledge multi-ontology framework to enrich ontology knowledge. Furthermore the ontology-based requirements elicitation approach can be obtained according to the above framework. The requirements error pattern can be weaved into the task ontology by aspect-oriented method. Finally, a case study is presented to validate the effectiveness of the proposed requirements elicitation approach. The results show that the number of errors in requirements document obtained by the proposed approach is less than the comparative approach. It shows that the proposed approach is effective.
Author Names:
Xuan Hu*, BeiHang University Yichen Wang, Beihang University Liu Bin, Minyan Lu, BeiHang University
Author Emails :
happydoghx AT dse.buaa.edu.cn wangyichen AT buaa.edu.cn liubin AT buaa.edu.cn lmy AT buaa.edu.cn
