Fake news diffusion is a primary driver of misinformation. Analyzing deliberately false and misleading content is tough because social media platforms make it incredibly easy to create and spread huge amounts of information quickly. The intricate dynamics of fake news propagation demand the availability of ready-to-use frameworks for its analysis. This paper explores the automatic topic identification component of SPREADSHOT, a graph-based method designed to analyze fake news dissemination by examining two key factors: spreaders and topics. When it comes to news content, fake news frequently revolves around rapidly evolving topics due to its strong connection to current events. Consequently, topic modeling has gained significant traction for analyzing news articles. In our analysis, we explore two distinct topic modeling techniques: Latent Dirichlet Allocation (LDA) and BERTopic. While both offer valuable insights, we carefully justify which of these two techniques is best suited for integration into the SPREADSHOT framework for topic modeling.

Current trends in network intrusion detection systems (NIDS) capitalize on the extraction of features from network traffic and the use of up-to-date machine and deep learning techniques to infer a detection model; in consequence, NIDS can be vulnerable to adversarial attacks. Differently from the plethora of contributions that apply (and misuse) feature-level attacks envisioned in application domains far from NIDS, this paper proposes a novel approach to adversarial attacks, which consists in a realistic problem-space perturbation of the network traffic. The perturbation is achieved through a traffic control utility. Experiments are based on normal and Denial of Service traffic in both legitimate and adversarial conditions, and the application of four popular techniques to learn the NIDS models. The results highlight the transferability of the adversarial examples generated by the proposed problem-space attack as well as the effectiveness at inducing traffic misclassifications across the NIDS models obtained.

Intrusion detection systems (IDS) play a key role to assure security properties of modern computer networks. IDS are often based on machine and deep learning techniques; as such, IDS are vulnerable to various forms of adversarial attacks. This paper presents an initial case study on the robustness of machine learning for network intrusion detection against adversarial attacks. Experiments are based on a recent fix of the widely-used CICIDS2017 benchmark dataset, two well-known machine learning techniques for intrusion detection (i.e., deep autoencoders and decision trees), and the virtual adversarial method (VAM) to generate the adversarial examples. Based on the data and experiments at hand, the results provide many interesting findings on the robustness of the IDS models assessed. The autoencoder-based IDS is more robust to evasion rather than overstimulation. On the contrary, the decision tree is vulnerable to evasion; moreover, changes to the learning parameters can strongly affect the robustness of the decision tree against the VAM attack.

Security of Internet of Things (IoT) devices and networks is a primary concern. Many intrusion detection systems (IDS) proposals in the IoT leverage machine and deep learning algorithms to learn models that can be used to discriminate normal behaviors from intrusions. Due to the dynamicity and scale of modern IoT networks, it is hard to learn and maintain one separate IDS model per device; on the other hand, the Cloud-Edge-IoT architecture allows learning a single IDS model (instead of many separate models). This paper compares two paradigms, i.e., traditional and federated, to learn a single IDS model atop the traffic of different IoT devices. The former assumes the availability of an all-in-one training dataset at a unique learning node; the latter aggregates the outcomes of independent learning procedures executed on individual training datasets hosted by different nodes. The experiments are done with a well-established public benchmark of nine IoT devices and the use of deep autoencoders. In the experiment and dataset at hand, federated learning lead to an increase of the false positive rate of six devices compared to the traditional scenario. Such an increase was balanced by a narrower variability of the false positive rate across all the devices and a mitigation of potential overfitting.

Detecting attacks to Cyber-Physical Systems (CPSs) is of utmost importance, due to their increasingly frequent use in many critical assets. Intrusion detection in CPSs and other domains, such as the Internet of Things, is often addressed through machine and deep learning. However, many existing proposals tend to favor the application of complex detection models over the usability in real-world operations. This paper presents CPS-GUARD, a novel intrusion detection approach based on a single semi-supervised autoencoder and a technique to set the threshold used to discriminate normal operations from attacks. The technique is outlier-aware, in that it relies on outlier detection to mitigate inherent imperfections of the training data. CPS-GUARD is evaluated by means of direct experiments with normal and intrusion data points pertaining to individual sensing devices, an HTTP server and four full-fledged systems, including CPSs. Experiments are based on a wide spectrum of attacks available in six state-of-the-art datasets. The intrusion detection results of CPS-GUARD are within 0.949-1.000 recall, 0.961-0.999 precision and 0.006-0.027 false positive rate depending on the specific system. The results are competitive with other existing intrusion detection methods. The evaluation is complemented by a comparative study on alternative threshold selection and outlier detection techniques.

In the past years Internet of Things (IoT) has received increasing attention by academia and industry due to the potential use in several human activities; however, IoT devices are vulnerable to various types of attacks. Many existing intrusion detection proposals in the IoT leverage complex machine learning architectures, which may provide one separate model per device or per attack. These solutions are not suited to the dynamicity and scale of modern IoT environments. This paper proposes an initial analysis of the problem in the context of deep autoencoders and the detection of botnet attacks. Our findings, obtained by means of the N-BaIoT dataset, indicate that it is relatively easy to achieve impressive detection results by training-testing separate and minimal deep autoenconders on the top of the data individual IoT devices. More important, our all-in-one deep autoencoding proposal, which consists in training a single model with the benign traffic collected from different IoT devices, allows to preserve the overall detection performance obtained through separate autoencoders. The all-in-one model can pave the way for more scalable intrusion detection solutions in the context of IoT.

The availability of ready-to-use public security datasets is fostering measurement-driven research by a wide community of academics and practitioners. Recent trends in this area put forth a substantial body of literature on anomaly and attack detection on the top of public labelled datasets. Much of this literature blindly reuses existing datasets by overlooking the cybersecurity facets of the network traffic therein, in terms of its real impact on service availability and performance of operations.

Applications executed in the cloud can exploit its elasticity features, varying dynamically the amount of leased resources so as to adapt to load variations and to guarantee quality of service. As auto-scaling has implications on execution costs, making optimal scaling choices is of paramount importance. This paper presents an analysis method based on offline benchmarking and simple models that allows to evaluate performance indexes useful to define scaling policies to be used by auto-scalers. The proposed approach relies on a fixed set of benchmarks, to be executed off-line and a set of models that enable prediction of the same performance indexes under different workload conditions, enabling the analyst to perform parameter analysis when defining an auto-scaling policy.

The International Conference on Enabling Technologies: Infrastructure for Collaborative Enterprises WETICE is an international forum for state-of the-art research in enabling technologies for collaboration. The 28th WETICE edition takes place on June 12-14, 2019 in Capri (Napoli), Italy and it is made of eleven scientific tracks.

Applications executed in the cloud can exploit its elasticity features, varying dynamically the amount of leased resources so as to adapt to load variations and to guarantee good quality of service. As auto scaling has severe implications on execution costs, making optimal scaling choices is of paramount importance. This paper presents an analysis method based on off-line benchmarking that allows to define scaling policies to be used by auto-scalers. The indexes obtained by benchmarking multiple deployment configurations can be used on-line, to scale the application making a trade-off between cost and user-perceived performance.

The development of cloud applications raises several security concerns due to the lack of control over involved resources. Security testing is fundamental to identify the existing security issues and is particularly powerful when carried out by means of penetration testing techniques. Unfortunately, penetration testing requires a deep knowledge of the possible attacks and of the available hacking tools and is very energy demanding. In this paper, we present a methodology that allows to easily carry out a coarse-grained security evaluation of a cloud application by automating the set-up and execution of penetration tests. The methodology relies on the knowledge of the application architecture and on the availability of a catalogue including security-related data collected from multiple sources and properly correlated.

The development of applications based on services offered by different, not conscious, providers, is expected to be growing in the next years. In order to offer effectively multicloud applications, many challenges still need to be faced. At this aim, the MUSA framework provides a DevOps approach to develop multi-cloud applications with desired Security Service Level Agreements (SLAs). This paper describes the MUSA Deployer models, which help developers to express their security requirements, and a Deployer tool that automatically provides cloud security services to offer Security SLAs.

The extensive use of cloud services by both individual users and organizations induces several security risks. The risk perception is higher when Cloud Service Providers (CSPs) do not clearly state their security policies and/or when such policies do not directly match user-defined requirements. Security-oriented Service Level Agreements (Security SLAs) represent a fundamental means to encourage the adoption of cloud services in contexts where security is mandatory. Nevertheless, despite the number of existing initiatives aimed at formalizing Security SLAs and at representing security guarantees by taking into account both customers' and providers' perspectives, they are far from being commonly adopted in practice by CSPs, due to the difficulty in automatically enforcing and monitoring the security capabilities agreed with customers. In this paper we illustrate, through a case study, a methodology to set-up a catalogue of security capabilities that can be offered as-a-service, on top of which specific guarantees can be specified through a Security SLA. Such a methodology, which explicitly takes into account the constraints behind the definition of formal guarantees related to security, is meant to serve as a guideline for providers willing to offer for their services specific security features that can be monitored and assessed by customers during operation.

In the cloud computing context, Service Level Agreements (SLAs) are contracts between Cloud Service Providers (CSPs) and Cloud Service Customers (CSCs), stating the guaranteed quality level of the services offered by CSPs. Existing cloud SLAs focus only on few service terms, completely ignoring all security related aspects. They are often reported in a way that is hardly understandable for customers. Moreover, they offer guarantees uniform for all offered services and all customers, regardless of particular service characteristics or customers specific needs. This paper presents a framework that enables the adoption of a per-service SLA model, by supporting the automatic implementation of cloud Security SLAs tailored to the needs of each customer for specific service instances. In particular, the process and the software architecture for per-service SLA implementation are shown. A case study application demonstrates the feasibility and effectiveness of the proposed solution.

In cloud computing, possible risks linked to availability, performance and security can be mitigated by the adoption of Service Level Agreements (SLAs) formally agreed upon by cloud service providers and their users. This paper presents the design of services for the management of cloud-oriented SLAs that hinge on the use of a REST-based API. Such services can be easily integrated into existing cloud applications, platforms and infrastructures, in order to support SLA-based cloud services delivery. After a discussion on the SLA life-cycle, an agreement protocol state diagram is introduced. It takes explicitly into account negotiation, remediation and renegotiation issues, is compliant with all the active standards, and is compatible with the WS-Agreement standard. The requirement analysis and the design of a solution able to support the proposed SLA protocol is presented, introducing the REST API used. This API aims at being the basis for a framework to build SLA-based applications.

Summary The writing of efficient parallel code has always been a tedious and time‐consuming process. However, the performance prediction prototype tools devised in the last decade come of age now, thanks to the availability of the almost unlimited computing power of clouds. This paper presents the practical use of mJADES, a novel environment for running multiple concurrent simulations in the cloud, to predict the performance of parallel code in multiple working conditions at once. After an introduction on mJADES and its operational aspects, the construction of parallel code performance prediction models will be dealt with. The models and the results obtained for a simple but complete and meaningful case study will be presented, discussing the accuracy obtained by simulation and the time required for performing the whole set of simulations necessary to characterize the program behavior. Copyright © 2013 John Wiley & Sons, Ltd.

Cloud computing is an emerging paradigm, recently widely adopted in distributed and business computing. Even if it is very attractive, due to its business model (pay-per-use) and its flexibility (self-service on demand approach), one of the main limits for its adoption is the perception of loss of security and control over resources that are dynamically acquired in the cloud and that reside on remote providers. Moreover, security mechanisms are usually integrated into system architectures, and are not offered to users in a way that enables customization and is easy to use. As a consequence, as far as security is concerned, cloud customers are usually tied to a limited set of offerings made available by providers, often without real grants about the way in which such mechanisms are actually implemented and enforced. This paper deals with the architecture underlying the SPECS platform, which aims at offering security features by an as-a-service approach, using Service Level Agreements as a mean for clear statement between customers and providers to define mutual rights and constraints. The goal is to show the main requirements of such platform and to present the global architecture, in terms of components and their interactions, dedicated to negotiate, to monitor and to enforce the security mechanisms to be applied over existing cloud providers.

The potential of cloud computing is still underutilized in the scientific computing field. Even if clouds probably are not fit for high-end HPC applications, they could be profitably used to bring the power of economic and scalable parallel computing to the masses. But this requires simple and friendly development environments, able to exploit cloud scalability and to provide fault tolerance. This paper presents a framework built on the top of a cloud-aware platform (mOSAIC) for the development of bag-of-tasks scientific applications.

The integration of sensing infrastructures into the Cloud gives a number of advantages in providing sensor data as a service over the Internet. Many solutions are now available in the literature, and most of them focus on modeling sensor networks as part of the infrastructure to be offered as a service (IaaS), directly managed by means of the Cloud tools that provide resource virtualization. We propose a different approach: sensor networks are modeled as providers that offer their resources to a Cloud application that runs independently from Cloud providers. Being offered as a Service, any user can negotiate with the provider his desired requirements in terms of operational parameters and non-functional features (i.e. security, dependability, etc). In particular, we propose a SLA-based approach for the specification and management of usage term guarantees related to the access and configuration of private sensor networks. To this end, a Cloud Sensing Brokering Platform is designed to illustrate the innovative way to integrate Cloud and Sensor Networks.

The cloud offers attractive options to migrate corporate applications, without any implication for the corporate security manager to manage or to secure physical resources. While this ease of migration is appealing, several security issues arise: can the validity of corporate legal compliance regulations still be ensured for remote data storage? How is it possible to assess the Cloud Service Provider (CSP) ability to meet corporate security requirements? Can one monitor and enforce the agreed cloud security levels? Unfortunately, no comprehensive solutions exist for these issues. In this context, we introduce a new approach, named SPECS. It aims to offer mechanisms to specify cloud security requirements and to assess the security features offered by CSPs, and to integrate the desired security services (e.g., credential and access management) into cloud services with a Security-as-a-Service approach. Furthermore, SPECS intends to provide systematic approaches to negotiate, to monitor and to enforce the security parameters specified in Service Level Agreements (SLA), to develop and to deploy security services that are cloud SLA-aware and are implemented as an open-source Platform-as-a-Service (PaaS). This paper introduces the main concepts of SPECS.

The simulation of complex systems is a time expensive and resource consuming task. Furthermore, the statistical validation of simulation models or the need to compare the system behavior in several different conditions require many simulation runs. A viable solution to speed up the process is to run multiple simulations in parallel on a distributed system, for example by means of a web interface. In this context, the cloud computing paradigm can be of help, offering almost unlimited computing resources that can be leased as a service. In this paper we will present the structure of a new simulation engine " in thecloud" (mJADES), which runs on the top of an ad-hoc federation of cloud providers and is designed to perform multiple concurrent simulations. Given a set of simulation tasks, mJADES is able automatically to acquire the computing resources needed from the cloud and to distribute the simulation runs to be executed.

Both cloud and GRID are computing paradigms for the large-scale management of distributed resources. Even if the first is usually oriented to transaction-based applications, and the latter to High Performance Computation, there is a lot of interest in their integration. This is typically obtained through the Infrastructure-as-a-Service cloud model, which is exploited in the GRID context to offer machine with full administration rights to users. In this paper the focus is on the security problems linked to the integration of cloud and GRID computing. It is proposed the adoption of identify federation between different security domains to manage the relationship between the user machines and the standard GRID infrastructure. This solution is experimented within PerfCloud, a cloud implementation that exploits an underlying GRID platform.

The highly distributed nature and the load sensitivity of Service Oriented Architectures (SOA) make it very difficult to guarantee performance requirements under rapidly-changing load conditions. This paper deals with the development of service oriented autonomic systems that are capable to optimize themselves using a feed forward approach, by exploiting automatically generated performance predictions. The MAWeS (MetaPL/HeSSE Autonomic Web Services) framework allows the development of self-tuning applications that proactively optimize themselves by simulating the execution environment. After a discussion on the possible design choices for the development of autonomic web services applications, a soft real-time test application is presented and the performance results obtained in a composite-service execution scenario are commented.

Abstract This paper describes a simulation‐based technique for the performance prediction of message‐passing applications on cluster systems by means of benchmark data. Given data measuring the performance of a target cluster in the form of standard benchmark results, along with the details of the chosen computing configuration, it is possible to build and to validate automatically a detailed simulation model. This makes it possible to predict off‐line, i.e. without resorting to the real hardware, the performance of fully developed or even of skeletal code. An XML‐based language (MetaPL) is adopted to describe the application behavior in the development stage. After a description of the approach and the illustration of the construction and validation of the simulation model, the paper presents a case study. Copyright © 2006 John Wiley & Sons, Ltd.

Service oriented architectures (SOA) are based on applications consisting of an aggregation of services with standard interface, offered on distributed hosts. The highly distributed nature and the load sensitivity of these architectures make it very difficult to guarantee performance requirements under rapidly-changing load conditions. This paper deals with the development of service oriented predictive autonomic systems that are capable to optimize themselves using a feedforward approach, by exploiting automatically generated performance predictions. The MAWeS (MetaPL/HeSSE autonomic Web services) framework allows the development of self-tuning applications that proactively optimize themselves by simulating the execution environment. An example of application development in MAWeS is thoroughly dealt with, showing the implementation of a system that exploits MAWeS services to choose dynamically among several different algorithms to meet response time constraints.

This paper addresses the problem of the measurement of efficiency in heterogeneous distributed computing systems. After a discussion on the unsuitability of the traditional notion of efficiency for such systems, a new efficiency metric (generalized efficiency) is introduced by finding the analytical expression of the maximum speedup that can be achieved for a given problem in a heterogeneous system. Generalized efficiency takes into account the heterogeneity of computing resources and is consistent with the definition commonly used for homogeneous systems. The main characteristics of this metric are illustrated, and the problem of performing practical measurements is dealt with thoroughly. Four case studies exploring the whole spectrum of heterogeneous computing systems and of possible classes of applications are presented. Finally, the proposal is compared to related work, and the utility and the limits of generalized efficiency are discussed. © 1998 John Wiley & Sons, Ltd.

Abstract This paper addresses the problem of monodimensional (1D) FFT parallel computation on constant‐valence multicomputers, i.e. on parallel systems made up of processing elements (PEs) which do not share memory and are connected to a bounded number of neighbours. After a qualitative analysis of several possible partitionings of the DIT FFT algorithm, a decomposition is introduced that has good scalability properties and makes it possible to use sections of sequential code based on the most common 1D‐FFT algorithms. If a computing architecture with indirect binary n‐cube interconnection network is used, the proposed decomposition guarantees strictly local communications and therefore requires no through‐routing support. These characteristics have a positive impact on software development and also on overall performance. Furthermore, thanks to a pipelined organization of the PEs, the resulting architecture has high potentialities for real‐time signal processing. As these useful features are obtained at the ‘expense’ of an uneven workload distribution, computing efficiency is relatively low but does not significantly change in a wide range of the number of processors. An implementation on a Transputer‐based system is presented along with the performance results obtained. Finally a simple analytical model of the architecture is shown, that allows the values of the main performance parameters to be obtained as a function of the number of processors used and of the elementary response times of the first stage of PEs.

Abstract The paper deals with the implementation of global time in multicomputer systems. After a formalization of the synchronization problem, techniques to estimate the synchronization delay and to compensate the drift error are proposed. Then SYNC_WAVE, a clock synchronization algorithm where the values of a reference clock are diffused in a wave‐like manner, is described. SYNC_WAVE has no provision for fault‐tolerance and is specially designed to introduce low CPU and communication overhead, in order to support performance analysis applications efficiently. An implementation of the devised algorithm in a transputer‐based system is presented, showing the accuracy results obtained. Finally SYNC_WAVE is compared to other synchronization algorithms and several of its possible applications are suggested.

Abstract Parallel programs are intrinsically non‐deterministic, and therefore the techniques of cyclical debugging that are commonly used for sequential programs are not suitable for parallel ones. This paper proposes a method to reproduce Occam program behaviour. Saving information on the timer values input by the program and the guards selected at run‐time on alternative commands allows program replay, i.e. it makes it possible to re‐execute the program deterministically with the same inputs following the same instruction path. This enables the software developer to use tools such as debuggers and intrusive monitors to help identify program faults. After discussing possible implementations of the proposed technique, IRD (an interactive replay debugger for Occam programs) is described. Finally, the use of the IRD in a sample debug session is presented as an example.

Abstract In a distributed system based on Transputer components there is one clock for each processing element, and the definition of the global system time requires the choice of a hardware or software synchronization method. This paper describes the RING_SYNC algorithm, based on a ring‐structured synchronization scheme. RING_SYNC has no provision for fault tolerance, but it introduces little overhead, thanks to the optimization of both the number of messages exchanged at sync time and the resynchronization frequency. The implementation of the algorithm together with the tests performed for measuring the synchronization error and their results are discussed extensively, and some typical applications are pointed out.