Federated learning (FL), a decentralized machine learning framework that allows edge devices (i.e., clients) to train a global model while preserving data/client privacy, has become increasingly popular recently. In FL, a shared global model is built by aggregating the updated parameters in a distributed manner. To incentivize data owners to participate in FL, it is essential for service providers to fairly evaluate the contribution of each data owner to the shared model during the learning process. To the best of our knowledge, most existing solutions are resource-demanding and usually run as an additional evaluation procedure. The latter produces an expensive computational cost for large data owners. In this paper, we present simple and effective FL solutions that show how the clients’ behavior can be evaluated during the training process with respect to reliability, and this is demonstrated for two existing FL models, Cluster Analysis-based Federated Learning (CA-FL) and Group-Personalized FL (GP-FL), respectively. In the former model, CA-FL, the frequency of each client to be selected as a cluster representative and in that way to be involved in the building of the shared model is assessed. This can eventually be considered as a measure of the respective client data reliability. In the latter model, GP-FL, we calculate how many times each client changes a cluster it belongs to during FL training, which can be interpreted as a measure of the client’s unstable behavior, i.e., it can be considered as not very reliable. We validate our FL approaches on three LEAF datasets and benchmark their performance to two baseline contribution evaluation approaches. The experimental results demonstrate that by applying the two FL models we are able to get robust evaluations of clients’ behavior during the training process. These evaluations can be used for further studying, comparing, understanding, and eventually predicting clients’ contributions to the shared global model. • New federated learning solutions for quantifying the contribution of each client to the overall global model according to its behavior. • The clients’ behavior can be evaluated during the training process concerning reliability. • Without incurring any considerable communication costs, our methods have shown robustness in evaluating each client’s behavior respectively contribution to the federated model and have the potential to be used for this purpose.

Automated analysis and specification of software requirements expressed in natural language is a challenge addressed by the research community and is becoming a reality thanks to the advances in Artificial Intelligence (AI) and Natural Language Processing (NLP) techniques. While the research community focuses mainly on generic software requirements or specialized solutions for security requirements, we find a gap in the automation of analysis and specification for requirements in the cloud computing domain and the automatic mapping of requirements on actual products offered in the cloud service market. In this research work, we propose AI-CRAS an AI-driven cloud service requirement analysis and specification methodology. The proposed method, which leverages state-of-the-art transformer-based large language model, has been implemented and validated in a real case. Experimental results demonstrate that the model performed well in binary and multi-label classification of requirements (achieving recall/F1-score of $0.96 / 0.92$ and $0.86 / 0.76$, respectively) and mapping requirements into actual cloud services.

Training of machine learning models in a Datacen-ter, with data originated from edge nodes, incurs high communication overheads and violates a user's privacy. These challenges may be tackled by employing Federated Learning (FL) machine learning technique to train a model across multiple decentralized edge devices (workers) using local data. In this paper, we explore an approach that identifies the most representative updates made by workers and those are only uploaded to the central server for reducing network communication costs. Based on this idea, we propose a FL model that can mitigate communication overheads via clustering analysis of the worker local updates. The Cluster Analysis-based Federated Learning (CA-FL) model is studied and evaluated in human activity recognition (HAR) datasets. Our evaluation results show the robustness of CA - FL in comparison with traditional FL in terms of accuracy and communication costs on both IID and non-IID cases.

Summary Containerization is a lightweight virtualization technology enabling the deployment and execution of distributed applications on cloud, edge/fog, and Internet‐of‐Things platforms. Container technologies are evolving at the speed of light, and there are many open research challenges. In this paper, an extensive literature review is presented that identifies the challenges related to the adoption of container technologies in High Performance Computing, Big Data analytics, and geo‐distributed (Edge, Fog, Internet‐of‐Things) applications. From our study, it emerges that performance, orchestration, and cyber‐security are the main issues. For each challenge, the state‐of‐the‐art solutions are then analyzed. Performance is related to the assessment of the performance footprint of containers and comparison with the footprint of virtual machines and bare metal deployments, the monitoring, the performance prediction, the I/O throughput improvement. Orchestration is related to the selection, the deployment, and the dynamic control of the configuration of multi‐container packaged applications on distributed platforms. The focus of this work is on run‐time adaptation. Cyber‐security is about container isolation, confidentiality of containerized data, and network security. From the analysis of 97 papers, it came out that the state‐of‐the‐art is more mature in the area of performance evaluation and run‐time adaptation rather than in security solutions. However, the main unsolved challenges are I/O throughput optimization, performance prediction, multilayer monitoring, isolation, and data confidentiality (at rest and in transit).

Summary NoSQL distributed databases are often used as Big Data platforms. To provide efficient resource sharing and cost effectiveness, such distributed databases typically run concurrently on a virtualized infrastructure that could be implemented using hypervisor‐based virtualization or container‐based virtualization. Hypervisor‐based virtualization is a mature technology but imposes overhead on CPU, networking, and disk. Recently, by sharing the operating system resources and simplifying the deployment of applications, container‐based virtualization is getting more popular. This article presents a performance comparison between multiple instances of VMware VMs and Docker containers running concurrently. Our workload models a real‐world Big Data Apache Cassandra application from Ericsson. As a baseline, we evaluated the performance of Cassandra when running on the nonvirtualized physical infrastructure. Our study shows that Docker has lower overhead compared with VMware; the performance on the container‐based infrastructure was as good as on the nonvirtualized. Our performance evaluations also show that running multiple instances of a Cassandra database concurrently affected the performance of read and write operations differently; for both VMware and Docker, the maximum number of read operations was reduced when we ran several instances concurrently, whereas the maximum number of write operations increased when we ran instances concurrently.

It has previously been assumed that the size of an IPv6 network would make it impossible to scan the network for vulnerable hosts. Recent work has shown this to be false, and several methods for scanning IPv6 networks have been suggested. However, most of these are based on external information like DNS, or pattern inference which requires large amounts of known IP addresses. In this paper, DeHCP, a novel approach based on delimiting IP ranges with closely clustered hosts, is presented and compared to three previously known scanning methods. The method is shown to work in an experimental setting with results comparable to that of the previously suggested methods, and is also shown to have the advantage of not being limited to a specific protocol or probing method. Finally we show that the scan can be executed across multiple VLANs.

Machine learning software accounts for a significant amount of energy consumed in data centers. These algorithms are usually optimized towards predictive performance, i.e. accuracy, and scalability. This is the case of data stream mining algorithms. Although these algorithms are adaptive to the incoming data, they have fixed parameters from the beginning of the execution. We have observed that having fixed parameters lead to unnecessary computations, thus making the algorithm energy inefficient. In this paper we present the nmin adaptation method for Hoeffding trees. This method adapts the value of the nmin parameter, which significantly affects the energy consumption of the algorithm. The method reduces unnecessary computations and memory accesses, thus reducing the energy, while the accuracy is only marginally affected. We experimentally compared VFDT (Very Fast Decision Tree, the first Hoeffding tree algorithm) and CVFDT (Concept-adapting VFDT) with the VFDT-nmin (VFDT with nmin adaptation). The results show that VFDT-nmin consumes up to 27% less energy than the standard VFDT, and up to 92% less energy than CVFDT, trading off a few percent of accuracy in a few datasets.

Today the cloud-desktop service, or Desktop-as-a-Service (DaaS), is massively replacing Virtual Desktop Infrastructures (VDI), as confirmed by the importance of players entering the DaaS market. In this paper we study the workload of a DaaS provider, analyzing three months of real traffic and resource usage. What emerges from the study, the first on the subject at the best of our knowledge, is that the workload on CPU and disk usage are long-tail distributed (lognormal, weibull and pare to) and that the length of working sessions is exponentially distributed. These results are extremely important for: the selection of the appropriate performance model to be used in capacity planning or run-time resource provisioning, the setup of workload generators, and the definition of heuristic policies for resource provisioning. The paper provides an accurate distribution fitting for all the workload features considered and discusses the implications of results on performance analysis.

Architecting software systems according to the service-oriented paradigm and designing runtime self-adaptable systems are two relevant research areas in today's software engineering. In this paper, we address issues that lie at the intersection of these two important fields. First, we present a characterization of the problem space of self-adaptation for service-oriented systems, thus providing a frame of reference where our and other approaches can be classified. Then, we present MOSES, a methodology and a software tool implementing it to support QoS-driven adaptation of a service-oriented system. It works in a specific region of the identified problem space, corresponding to the scenario where a service-oriented system architected as a composite service needs to sustain a traffic of requests generated by several users. MOSES integrates within a unified framework different adaptation mechanisms. In this way it achieves greater flexibility in facing various operating environments and the possibly conflicting QoS requirements of several concurrent users. Experimental results obtained with a prototype implementation of MOSES show the effectiveness of the proposed approach.

The complexity of cloud systems poses new infrastructure and application management challenges. One of the common goals of the research community, practitioners and vendors is to design self-adaptable solutions capable to react to unpredictable workload fluctuations and changing utility principles. This paper analyzes the problem from the perspective of an application service provider that uses a cloud infrastructure to achieve scalable provisioning of its services in the respect of QoS constraints. In the specific, we propose four architectural schemas for autonomic resource management of cloud-based systems. The proposed solutions, differing for the degree of control on the autonomic cycle phases, have been designed considering: (1) functional requirements dictated by the resource provisioning solutions proposed in literature; and (2) services currently offered by public IaaS providers. Finally, we propose an implementation of our solution based on Amazon EC2.

In this paper we introduce MORE-GIS, a simulator for the simulation of mobile nodes connected to wired and wireless network for the study of critical infrastructures interdependencies. MORE-GIS is based on agent based simulation approach and HLA architecture. Every node is an agent that represents a person. Agents are connected to Internet generating a certain amount of traffic and can move to a new geographic position. Movements are simulated using GIS data of roads and junctions. Every person selects the path using the Dijkstra's algorithm. Actually we use repast symphony framework for the agent based simulation and the mobility framework of OMNeT++ for the simulation of the wireless and wired communication network. They are synchronized and exchange information using the HLA architecture.

Runtime adaptation is recognized as a viable way for a service-oriented system to meet QoS requirements in its volatile operating environment. In this paper we propose a methodology to drive the adaptation of such a system, that integrates within a unified framework different adaptation mechanisms, to achieve a greater flexibility in facing different operating environments and the possibly conflicting QoS requirements of several concurrent users. To determine the most suitable adaptation action(s), the methodology is based on the formulation and solution of a linear programming problem, which is derived from a behavioral model of the system updated at runtime by a monitoring activity. Numerical experiments show the effectiveness of our approach. Besides the methodology, we also present a prototype tool that implements it.

Agent-based modeling and simulation (ABMS) is one of the more promising simulation techniques to study the interdependencies in critical infrastructures. Moreover, federated simulation has two relevant properties, simulation models reuse and expertise sharing, that could be exploited in a multi-sectorial field, such as critical infrastructure protection. In this paper we propose a new methodology which exploits the benefit of both ABMS and Federated simulation, to study interdependencies in critical infrastructures. First of all we discus advantages of federated agent-based modeling and difficulties in implementing a Federated ABMS framework. To demonstrate the relevance of our solution we propose an example driven approach that poses the attention on critical information infrastructure. We have also implemented a Federated ABMS framework, which federate Repast, an agent-based simulation engine and OMNeT++ an IT systems and communication networks modeling and simulation environment. A selection of simulation results shown how Federated ABMS could shed light on system interdependencies and how it helps in quantifying them.

In the service oriented paradigm applications are created as a composition of independently developed Web services. Since the same service may be offered by different providers with different non-functional Quality of Service (QoS) attributes, a selection process is needed to identify the constituent services for a given composite service that best meet the users QoS requirements. In this paper, we consider a broker that offers a composite service with multiple QoS classes to several users each generating a flow of requests over time. We propose a service selection scheme which optimizes the end-to-end aggregated QoS of all incoming flows of requests by means of a simple linear programming problem which scales as the number of users, request volumes and/or services grows. This approach differs from most of the current proposals which may not scale well since: a) requests, even from the same user, are handled independently from one another; and b) the selection process often requires the solution of an NP-hard problem.

One of the main issues in content-based publish/subscribe (CBPS) systems is how to dynamically determine groups of similar subscriptions to be adopted for exploiting efficient multicast techniques while guaranteeing at the same time the expressiveness of the subscription scheme. In this work, we propose a distributed mechanism which aims at satisfying important requirements of CBPS systems, that are: i) to guarantee the expressiveness of the subscription languages typical of the content-based paradigm, ii) to exploit efficient events dissemination, iii) to maintain the system scalability in terms of nodes and subscriptions, iv) to start an adaptive system reconfiguration despite new incoming subscriptions. One of the main feature of the proposed mechanism is the use of the system state knowledge sharing by system nodes, with the goal of limiting the system overhead in terms of computing, bandwidth and storage resources. Through a set of simulations we demonstrate the efficiency of the proposed solution.

In present wireless networks, mobile users frequently access continuous and session-oriented Internet services. During the handover, the management of session-related information introduces additional overheads and delays, due to context transfer procedures. Such delays may affect the QoS perceived by mobile users, making more difficult to realize seamless handover procedures. In this paper, we propose a framework to design a layer-2 trigger on the mobile node that intelligently activates the Context Transfer Protocol (CTP). Our proposed solution is based on methods to forecast the handoff time of the mobile node and the access router that will handoff the connection, and a model to estimate the time needed to complete the context transfer procedure. We show that the forecasting algorithm is stable, is able to effectively avoid the ping-pong effect, and converges both for simple and complex trajectories, typical of urban regions. We also evaluate the performance of CTP in the real case of a GSM network.

To ensure probabilistic guarantees on quality of service in content delivery networks (CDN), an access control support is needed that takes into account a proper differentiation of requests and performs session based decisions, managing different types of services and different service phases. In this paper we introduce a CDN architecture with access control capabilities at session aware access routers. We formulate a Markov modulated Poisson decision process for access control that captures the heterogeneity of multimedia services and the variable availability of resources due to the network congestions that characterize a non-dedicated network environment. The structural properties of the optimal solutions are studied and considered as the basis for the formulation of heuristics that perform close to the optimal policy.

The typical Web cluster architecture consists of replicated back-end and Web servers, and a network Web switch that routes client requests among the nodes. In this paper, we propose a new scheduling policy, namely client-awarepolicy (CAP), for Web switches operating at layer-7 of the OSI protocol stack. Its goal is to improve load sharing in Web clusters that provide multiple services such as static, dynamic and secure information. CAP classies the clientrequestson the basis of their expected impact on main server resources, that is, network interface, CPU, disk. At run-time, CAP schedules client requests reaching the Web cluster with the goal of sharing all classes of services among the server nodes. We demonstrate through a large set of simulations and some prototype experiments that dispatching policies aiming to improve locality in server caches give best results for Web publishing sites providing static information and some simple database searches. When we consider Web sites providing also dynamic and secure services, CAP is more eective than state-of-the-art layer-7 Web switch policies. The proposed client-aware algorithm is also more robust than server-aware policies whose performance depends on optimal tuning of system parameters, veryhardtoachieveina highly dynamic system suchasaWeb site. Categories and Subject Descriptors C.2.4 [######## ############# ########]: Distributed Systems; C.4 [########### ## #######]: Design studies; H.3.5 [########### ####### ### #########]: Online Information Services|Web-based services General Terms Algorithms, Design, Performance Keywords Load balancing, Dispatching algorithms, Clusters Copyright is held by the author/owner. WWW10, May 1-5, 2001, Hong Kong. Copyright 2001 ACM 1-58113-348-0/01/0005 ...$5.00. 1.

Cluster systems are leading architectures for building popular Web sites that have to guarantee scalable services when the number of accesses grows exponentially. The most common Web cluster systems consist of replicated back-end and Web servers and a Web switch that routes client requests among the nodes. We propose a new scheduling policy, namely Multi-Class Round Robin (MC-RR) for Web switches operating at layer 7 of the OSI protocol stack. Its goal is to improve load sharing in recent Web clusters that provide multiple services such as static and dynamic information. We demonstrate through a wide set of simulation experiments that other dispatching policies aiming to improve locality in server caches give best results for traditional Web publishing sites providing static information and some simple database searches. On the other hand, when we consider more recent Web sites providing highly dynamic services, MC-RR is much more effective than state-of-the-art Web switch policies. The proposed algorithm has the additional benefit of guaranteeing stable results because its performance does not depend on several parameters that are very hard to tune in highly variable Web systems.