In trust management systems, the trustor should be able to select a trustee candidate that has the maximum trustworthiness degree toward a specific goal and an amount of risk consistent with her/his risk acceptance degree. In this research, a novel computational trust model based on the principles of uncertainty theory is introduced. In the proposed model, trust is considered to be constructed of trustworthiness components. To calculate each of these trustworthiness components, empirical distributions of recommenders and trustor’s opinions about the existing trustworthiness and risk degrees of the trustee candidates are aggregated. In the decision making stage, the trustee candidate with the optimum trustworthiness and risk degrees is selected according to uncertain goal programming. Based on this method, trustworthiness and risk degrees of the trustee candidates are calculated according to the amount of negative and positive deviations from the optimal state. To verify the accuracy of the model’s behavior, a series of simulation scenarios are constructed. The results of these simulations demonstrate that the proposed model effectively selects the best trustee candidate according to parameters such as context, priorities of the trustworthiness components, trustor’s constraints and the trustworthiness and risk acceptance degrees. Finally, by comparing the model with other commonly used computational trust modeling approaches, it is shown that the proposed model has a lower mean absolute error (MAE) and produces more accurate results.

Abstract One of the main challenges of the security of cyber‐physical systems (CPSs) is the lack of an efficient approach to evaluate the impacts of attacks on physical processes and their probabilities of occurrence. This paper proposes a method for evaluating the security of CPSs. By using the proposed method, one can predict the attacker's preferences in attacking CPSs and study the dynamic behavior of systems under security attacks. To deal with uncertainties in attacker's decision making to conduct attacks, the attack tree structure is used and parameterized with suitable fuzzy data. In the next step, the fuzzy technique for order of preference by similarity to ideal solution method is used to evaluate the model and predict the behavior of attackers. Besides, the dynamic behavior of CPSs under attacks is investigated by using the system's process model. The output of the model is a relative estimation of the security level of system based on suitable security metrics, such as the probability of attack scenarios, the time that the process is able to operate after conducting attack before shutting down (time‐to‐shutdown), and security risks. We illustrate the effectiveness of the method by comparing it with another attack tree‐based method. Furthermore, we present two illustrative examples and estimate the defined quantitative security measures. Copyright © 2017 John Wiley & Sons, Ltd.

Abstract Observational determinism is a property that ensures the confidentiality in concurrent programs. It conveys that public variables are independent of private variables during the execution of programs, and the scheduling policy of threads. Different definitions for observational determinism have been proposed. On the other hand, observational determinism is not a standard property and it should be checked over two or more executions of a program. The self-composition approach allows comparing two different copies of a program using a single formula. In this paper, we propose a new specification for the observational determinism security property in linear temporal logic. We also present a general method to create the appropriate program model using the self-composition approach. Both the program model and the observational determinism property are encoded in embedded C codes in PROMELA using the SPIN model checker. The paper also discusses a method for the instrumentation of PROMELA code in order to encode the program model for specifying the observational determinism security property.

The goal of privacy preserving clustering (PPC) is to preserve the privacy of data during clustering analysis. Most of the existing PPC algorithms are based on heuristic notions without provable privacy. Differential privacy is the strong notion of privacy introduced to overcome this problem. Howev er, the lower degree of utility is the serious drawback of the techniques, which preserve differential privacy. In addition, high dimensionality of data is another drawback of the most existing PPC techniques, which leads to low efficiency of them. This paper proposes differential-based algorithms for PPC in horizontally and vertically distributed datasets. To overcome the above two drawbacks, we have used orthogonal discrete wavelet transforms (DWT) for obtaining perturbed data with both low data dimensionality and less noise addition. Our algorithms are implemented and experimented using some well-known datasets. The results show that the proposed algorithms guarantee an appropriate level of both utility and privacy of the published data.

In this paper, we investigate some important aspects of a new formalism for modelling and verification of hybrid dynamic systems (HDS), which is called multisingular hybrid Petri nets (MSHPNs). This new hybrid formalism is aimed to bridge the gap between hybrid automata (HA) and hybrid Petri nets (HPNs) by equipping the HPN model with the capabilities of HA to control the execution and firing of timed transitions. Practically, MSHPNs can be considered as the counterpart with the same expressive power as multisingular hybrid automata (MSHA). In order to analyse MSHPN models, a speed-based partitioning technique has been introduced in which the variable space is partitioned based on the balance of continuous places. In this paper, we formalize the notions of conflicts and conflict resolution and the challenging issue of speed computation. Then, we focus on considering a translation from a bounded MSHPN to a multisingular hybrid automaton that preserves the behavioural semantics of the original MSHPN in terms of weak timed bisimulation. The translation algorithm uses the speed-based partitioning method and obtains a speed-based partitioning hybrid automaton for a given bounded MSHPN. Model checking a timed property for an MSHPN amounts to model checking its equivalent property on the obtained speed-based partitioning hybrid automaton, thus MSHPN models can be analysed using the existing tools. The advantages of the proposed method are twofold: (1) hybrid systems can be described more succinctly and therefore more readably as MSHPNs, and (2) one can use the existing tools (like HYTECH) to analyse MSHPN models.

So far, several techniques have been proposed for privacy preserving clustering (PPC). Most of the existing techniques have been designed based on heuristic notions without provable privacy guarantees. ϵ -differential privacy is a strong notion of pr

An appropriate model of attacker behaviour is a key requirement for quantitative security evaluation. Motivated by the fact that attacker behaviour is affected by some social factors such as monetary costs and benefits rather than merely the technical aspects of the target system, we proposed an attack modelling approach based on a hierarchical and coloured extension of stochastic activity networks HCSANs. This approach is called HCSAN-based attack modelling. By using this approach, multistage attacks can be modelled following the attack tree paradigm. Also, attacker behaviour can be modelled as a strategic decision-making process that accounts for the following factors affecting the attacker's decisions: 1 the goals of attack; 2 the cost and risk associated with available strategies; and 3 the target system's possible responses. Furthermore, we put forward an analytic solution method to measure security attributes i.e. confidentiality, integrity and availability and estimated two important quantitative security measures, which are the mean time to security failure and attack success probability. Additionally, we introduce a parametric sensitivity analysis method, which can be used to determine the sensitivity of the evaluated measures to different model parameters and optimize the model accordingly. Finally, we demonstrated how this approach can be used for survivability enhancement of the system using a well-known risk assessment process. Copyright © 2013 John Wiley & Sons, Ltd.

The scanning method of computer worms is an important modeling parameter. Active worms use different methods to scan the network for selecting their victim hosts. Topology-aware active worms scan the network by using the information of the network topology. This property, makes P2P network a suitable place for propagation of this kind of worms. Regarding the instability of network topology, especially in P2P networks, investigating the effect of the join and leave of hosts is completely necessary. In this paper, we study the propagation of topology-aware active worms based on SIR epidemiological model considering the join and leave of hosts. To have a better look at the effect of the join and leave of hosts on the propagation of active worms, the size of epidemic and infectious ratio parameters have been used. The results of this modeling show that the join and leave of hosts have considerable impact on the size of epidemic and the propagation performance of topology-aware active worms.

Today, the propagation of active worms in peer-to-peer (P2P) networks is a very serious issue. Therefore, studying the propagation behavior of active worms and modeling their proliferation is quite useful to employ an effective and low-cost defense strategy to reduce the rate of their propagation. The topology-aware worms use the information related to the topology and the network structure of their victims to propagate in the network. Recently, the logic matrix has been utilized to model the effect of worm propagation. In this paper, we use this kind of matrices to model the propagation of topology-aware active worms with emphasis on the impacts of the join and leave of hosts on the propagation ratio. Using a simulation model, we have studied and compared the impacts of the join and leave of hosts with the situations without them. The simulation results show the considerable impacts of join and leave on the propagation of worms.

Despite the benefits of data mining in a wide range of applications, this technique has raised some issues related to the privacy and security of individuals. Due to these issues, data owners may prevent to share their sensitive information with data miners. On the other hand, in distributed enviro nments, other issues related to the distribution of data will raise, which will make the preservation of privacy more challengeable. To solve these problems, different privacy preserving data mining (PPDM) techniques have been introduced. In this paper, a new privacy preserving clustering (PPC) technique for horizontally and vertically distributed datasets is proposed. The proposed technique uses Haar wavelet transforms (HWT) and scaling data perturbation (SDP) to achieve both data hiding and data reduction for protecting private numerical attribute values in distributed datasets. The results of our evaluations demonstrated that the proposed technique provides a proper degree of privacy and quality of clustering for distributed datasets and also runs fast. Our experiments have also shown that the proposed technique provides better privacy and clustering results comparing to the other existing privacy preserving clustering techniques applicable to distributed datasets. The proposed algorithms and the results of their experimental evaluations using different datasets are presented in this paper.

Security is the main concern and a challenging problem of Web services. In the recent years, XML-sensitive security appliances, such as XML firewalls or Web service firewalls, have been introduced to protect services. However, attackers can still compromise web services and do their malicious actions. Intrusion detection systems (IDS) are appropriate for defence in depth; and sit behind of firewalls in the security structure of an enterprise. However, network IDSs fail to detect attacks in Web service layer. In this paper, we propose an intrusion detection system for web services (WS-IDS), to detect malicious behaviors of the requesters of a typical web service. This idea is motivated by considering the inability of the existing IDSs to detect the attacks in web service layer. WS-IDS can be used in addition to other security appliances for web services, such as web service/XML firewalls.

The aim has been to propose a fraud detection system with capabilities of minimizing false alarms. In this paper we introduce a technique, which uses a hybrid fraud scoring and spike detection technique in streaming data over time and space. The technique itself differentiates normal, fraud and ano malous links, and increases the suspicion of fraud links with a dynamic global black list. Also, it mitigates the suspicion of normal links with a dynamic global white list. In addition, this technique uses spike detection technique to highlight the sudden and sharp rises in data, which can be indicative of abuse. The purpose is to derive two accurate suspicion scores for all incoming new examples in real-time. Results on mining several thousand credit application data demonstrate that the proposed technique reduces false alarm rates while maintaining a reasonable hit rate. In addition, new insights have been observed from the relationships between examples. The proposed technique takes the advantages of anomaly detection and supervised techniques. However by employing the spike detection technique, the false alarm rate is decreased. By this novel integration of techniques, the proposed technique is able to foil fraudsters' attempts, which continuously morph their styles to avoid to be detected. The results of the experiments to demonstrate the benefits of the technique are also presented in this paper.

The number of trademarks is growing rapidly, which makes searching among a large database of trademarks very difficult. The aim has been to develop an automated system to facilitate searching, comparing and retrieving trademarks. In this paper, such a software system is introduced. This software has three important functionalities: (1) annotation and storing trademarks it in a database, (2) text-based search and retrieval of trademarks based on annotations, and (3) content-based search and retrieval based on trademark images. In the proposed software system, we have implemented two integrated methods for retrieval of trademarks: text-based image retrieval (TBIR) and content-based image retrieval (CBIR) using several effective techniques. We introduce the techniques, functionalities and features of the software in this paper.

The aim is to develop a suitable method for quantifying security. We use stochastic modeling techniques for this purpose. An intrusion process is considered as a series of elementary attack phases and at each phase the interactions between the attacker and the system are analyzed rigorously. It is assumed that a typical attacker needs some time to perform an elementary attack phase. On the other hand, it is assumed that the attacker may be detected by the system and thus the overall intrusion process is interrupted. The attacker skill level and the system's abilities are characterized by the uniform distribution functions assigned to the transitions of the model. The underlying stochastic model is recognized as a semi-Markov chain. For security analysis, some valid assumptions about intrusion process are considered. Also, two quantitative security measures are defined and evaluated based on the model. The proposed method is demonstrated by modeling a complicated attack process and evaluating the desired security measures.

In data stream management systems (DSMSs), data do not appear in the form of persistent relations, but rather arrives in multiple, continuous, rapid, time-varying streams. Achieving a good performance in these systems is still the main challenge. Minimizing run-time memory usage and response time are the most important performance issues. Choosing a better scheduling algorithm implies a better performance. Variety of schedulers and customized quality of service (QoS) metrics (related to the amount of user's satisfaction), motivated us to find an approach for choosing the best scheduler per case. In this paper, a new static periodic scheduler called meta-scheduler is proposed. Our concentration is on non-real-time DSMSs dealing with semi-regular streams (not too bursty) using complex QoS metrics. We have used coloured Petri net models to choose the best scheduling algorithm for each period regarding complex QoS metrics and varying system statistics. We have studied our scheduler in the context of our new DSMS prototype. Finally, we showed that how meta-scheduler outperforms simple schedulers when a user defines a complex QoS metric.

In this paper, we introduce an object-oriented modeling framework called UMF for modeling with graph-based formalisms, such as Petri nets and related models. UMF is based on a four layered model for defining a concrete model, including meta-formalism, formalism, model class and models. We have defined these concepts formally and have provided a sample model for showing how to use them in UMF. We have used the object constraint language (OCL) for precise definition of formalism in UMF. There is also an innovative model composition approach in UMF, which is more flexible than other existing approaches for composing models of variant types.

Any system during its lifecycle could be subject to internal or external accidental or malicious threats. Therefore attention to system security is very important nowadays. Security brings in concerns for availability, in addition to confidentiality and integrity. Many security assessment methodologies like ITSEC, CC and etc were used up to now, but most of them have some limitations for being used in design phase of systems. In this paper, simulation of a network system for quantitative security evaluation (QSE) based on discrete-event simulation (DES) by SimEvents is presented. First, the system in normal state is simulated then an attacker is modeled as a client by means of zombies attacks to the system. Subsequently the availability of system begins to decrease. Finally the system cannot respond to the requests. After this simulation, the availability of system as an important security measure in any moment of simulation time is measured. At the end, a case study of distributed denial of service (DDoS) simulation is presented and the availability measure of the system is evaluated.

The most serious problem in the area of quantitative security evaluation is modeling of hacker's behavior. Because of the intelligent and complicated mental aspects of hackers, there are many challenges to model their behavior. Recently, there have been some efforts to use game theory for predicting hacker's behavior. However, it is necessary to revise the proposed approaches if there is a society of hackers with significant diversity in their behaviors. In this paper, we have examined our newly introduced approach to extend the basic ideas of using game theory to predict transition rates in stochastic models. The proposed method categorizes the society of hackers based on two main criteria used widely in hacker classification: motivations and skills. Markov chains are used to model the system. Based on the preferences of each class of hackers and the distribution of skills in each class, the transition rates between the states are computed. The resulting Markov chains can be solved to obtain the corresponding security measures of the system. We have explored some of the applications of the method and have shown that the method facilitates the study of relationships between important factors of hackers/defenders societies and different security measures of the system.

This paper proposes a high-performance network monitoring software architecture. The proposed architecture, named DashNMon, is able to employ multi-core CPUs in an efficient and scalable manner. Multi-core awareness is a distinguished property of this architecture. In spite of most existing cluster-based solutions, DashNMon can be used with common-off-the-shelf (COTS) multi-core CPUs. DashNMon is based on DashCap high-performance packet capture and transmission software solution, which we have recently introduced. Using the proposed architecture, it is possible to design and implement high-performance multi-threaded NIDSs or application-layer firewalls, completely in the user space and with better utilization of computational resources of multi-processor/multi-core systems. In this paper, after a brief overview of DashCap, we introduce the scalable software architecture of DashNMon and the results of the experiments carried out using a prototype web filter to benchmark its performance and scalability.

Web services are software components defined by WSDL, registered by UDDI and invoked by SOAP protocols. The port used by web services and SOAP is not typically blocked by conventional firewalls. Therefore, a new type of firewall named web service firewall or XML firewall is required. There are a number of commercial web service firewall products. Several academic projects have also been done on web service firewalls. This category of web service firewalls are briefly introduced in this paper and their drawbacks are mentioned. We have proposed a web service firewall architecture that supports authentication and authorization mechanisms. It also provides prevention of SOAP-based attacks. In this paper, we will present the proposed architectural design for a web service firewall. A formal model for the access control of the proposed architecture using coloured Petri nets (CPNs) is also presented. The CPN model is used for the analysis of the proposed architectural design. The model can also be served as a high-level design for implementation of the web service firewall.

Today, e-mail is one of the most popular communication facilities. E-mail list service (ELS) is a useful facility for scientific, commercial and social user groups. The main concern about ELS is how to manage and control the access of different users. To secure ELS by using security protocols and encryption/decryption mechanisms as well as user authentication, the computational overhead of servers will increase considerably. In addition, these methods do not offer adequate mechanism for entrance and subscription of users. In this paper we propose a trust-based scheme for managing users and controlling their access, with a minimum overhead on server. This scheme also manages the behaviour of users dynamically and establishes a two way connection between them.

Parallel programming and distributed programming are two solutions for scientific applications to provide high performance and fast response time in parallel systems and distributed systems. Parallel and distributed systems must provide inter process communication (IPC) mechanisms like message passing mechanism as underlying platforms to enable communication between local and especially geographically dispersed and physically distributed processes. Communication overhead is the major problem in these systems and there are a lot of efforts to develop more efficient message passing mechanisms or to improve the network communication speed. This paper provides hard evidence that an efficient implementation of message passing mechanism on multi-computers reduces the execution time of a molecular dynamics code. A well-known program for macromolecular dynamics and mechanics called CHARMm is executed on a networked cluster. The performance of CHARMm is measured with two distributed implementations of message passing, namely a kernel-level implementation called DIPC2006 and a renowned library level implementation called MPI. It is shown that the performance of CHARMm on a DIPC2006 configured cluster is by far better than its performance on an optimized MPI configured similar cluster. Even ignoring the favorable points of kernel-level implementations, like safety, privilege, reliability, and primitiveness, the insight is twofold. Scientists are nowadays faced with more computational complexity and look for more efficient systems and mechanisms. Efficient distributed IPC mechanisms have direct effect on running scientistspsila simulations faster, and computer engineers may try harder to develop more efficient distributed implementations of IPC.

Model refactoring is the process of transformation of model, which changes its internal structure for improving the qualitative features. Model refactoring should preserve the behavior of the model. Therefore, it is required to use a method to direct the process of model refactoring. It is also needed to use a formal method for accurate and unambiguous description of the model and its behavior. In this paper, we have focused on behavior preservation in refactoring UML diagrams. For this purpose, we have introduced control-flow diagram (CFD), which directs the process of model refactoring. We have also used refinement object-oriented language (ROOL), to prove that the refactoring technique we have used will preserve the behavior of models.

Call centers, or their contemporary successors, contact centers, are the preferred and prevalent way for many companies to communicate with their customers. The call center industry is thus vast and rapidly expanding in terms of both workforce and economic scope. It enjoys an annual growth rate of 20% and, overall, more than half of the business transactions are conducted over the phone. The design of such a system, and the management of its performance, surely must be based on sound scientific principles. Our goal in this paper is to survey the structure and technology of call center systems, point out some available special modeling and analysis tools, propose a high-level generalized stochastic Petri net (GSPN) model, and evaluate the model, aiming to minimize the workload of agents (servers) in order to provide self-serve.

We have recently introduced a high-level extension for stochastic activity networks (SANs) called coloured stochastic activity networks (CSANs). CSANs have several distinguishing properties, which make them quite appropriate for modeling and evaluation of software performance and dependability. CSANs have introduced a construct called coloured place for data manipulation. A coloured place holds a list of tokens of a userdefined token type. CSAN models can be evaluated by state space analysis techniques or discrete-event simulation. However, their state spaces will become very large, even for a small CSAN model. For efficient evaluation of these models by state space analysis methods, we will introduce measure-adaptive state space analysis process in this paper. Based on this method, it is possible to construct high-level CSAN models. However, for efficient evaluation, it is possible to generate and analyze a reduced state space based on user-specified performance or dependability measures.

Stochastic activity networks (SANs) are a stochastic generalization of Petri nets. SAN models have been used to evaluate a wide range of systems and are supported by several modeling tools. We have introduced object stochastic activity networks (OSANs) to overcome some restrictions of these models. OSANs integrate the concepts of object-orientation into SAN models. Elements of OSANs and their submodels are defined as classes. OSANs are more appropriate that most other objectoriented or high-level extensions of Petri nets for application on software systems. In this paper, we will present the definitions, behavior and an example of OSAN models. The objectorientation of OSANs and the flexibility of having functions for activities, make these models more appropriate than other extensions of Petri nets for modeling and evaluation of software systems.