With the recent dramatic increase in performance of artificial intelligence and related computing systems, together with advanced sensing, connectivity, and technological platforms, autonomous systems are poised to enter many application domains such as transportation and manufacturing. As autonomy increases, the risks of cybersecurity threats are equally rising, requiring the development of sophisticated methods on all layers of autonomous systems architectures. Therefore, this paper systematically introduces cybersecurity challenges ranging from the physical layer to the system of systems layer defining the collaboration of autonomous systems. Without loss of generality, autonomous vehicles are used to highlight current developments, illustrating which efforts are necessary to achieve secure and safe autonomous systems. Our discussions are comprehensively highlighting which research domains require further investigation and offer promising opportunities to contribute to mitigating cybersecurity challenges of autonomous systems.

In schedulability analysis, the timing of all tasks of a real-time system is verified by finding the worst-case behavior. Two well-studied methods in this field are the demand bound test and the real-time calculus. However, the former is only applicable to specific task models whereas the latter does not formalize concrete task models of complex job patterns. This work presents a new approach to formally describe and analyze the worst-case of any complex job pattern. The approach consists of a task model that reduces all kinds of job patterns to a vector space of jobs. Furthermore, the worst-case analysis searches for local maxima to find the worst-case of any job pattern by differentiating any cumulative function of the real-time calculus. Therefore, the analysis in this work implies an algorithm to compute request as well as demand bounds by construction. This formal approach allows the integration of mathematical results from real-time calculus into real-time scheduling theory. In fact, this is, to our knowledge, the first method to compute a demand bound function with arbitrary deadlines by a min-plus deconvolution in the real-time calculus. This now allows the analysis of complex task models as the generalized multiframe model in the real-time calculus.

In this paper, we describe a novel approach for spatial crowd analytics using WiFi signals emitted by smartphones following a non-intrusive approach. Our approach works with the use of WiFi sensors that we call surveyors. The surveyors, after being installed in an area, can be used to precisely track locations of visitors by trilaterating the source of the signals using distances estimated via the Received Signal Strength Intensity (RSSI). The evaluation of this system is performed in a real world scenario, in a shopping center located in Germany. Our system is evaluated on three different levels. The first level is the footfall estimation, which describes the number of people who walked in the shopping center. The second is the location tracking accuracy, and the third is the store engagement accuracy, which describes the number of people enter or stayed longer that specific time inside or in-front of a store. Overall, we have achieved a median 91.8% similarity on estimated footfall with an existing and reliable footfall counting system, we obtained 4.52m median tracking error, and 70.62% average store engagement accuracy.

In this paper, we consider Big Data applications for connected vehicles, where vehicle data is sent several times per second, permitting fine-grained and near real-time analysis of vehicle status and operating behavior. We focus on typical streaming applications and present implementations in Apache Spark and Apache Flink in different architectures and analyze several aspects of Big Data architectures for connected vehicles. First, we aim to show the potential of fine-grained analysis in terms of time resolution and level of detail for vehicular services. For instance, we show how to compute energy efficiency, comparing used energy vs needed energy, on the level of seconds. Secondly, we compare the architecture challenges in vehicular systems, including in-vehicle processing as well as cloud and edge processing. For these, we compare different approaches and architecture solutions. Third, we show that the performance and scalability of our different Big Data processing options differ significantly different for our use case.

As software architecture practice relies more and more on runtime data to inform decisions in continuous experimentation and self-adaptation, it is increasingly important to consider the quality of the data used as input to the different decision-making and prediction algorithms. One issue in data-driven decisions is that real-life data coming from running systems can contain invalid or wrong values which can bias the result of data analysis. Data-driven decision-making should therefore comprise detection and handling of data anomalies as an integral part of the process. However, currently, anomaly detection is either absent in runtime decision-making approaches for continuous experimentation and self-adaptation or difficult to tailor to domain-specific needs. In this paper, we contribute by proposing a framework that simplifies the detection of data anomalies in timeseries-outputs of running systems. The framework is generic, since it can be employed in different domains, and tunable, since it uses expert user input in tailoring anomaly detection to the needs and assumptions of each domain. We evaluate the feasibility of the framework by successfully applying it to detecting anomalies in a real-life timeseries dataset from the traffic domain.

In this paper, we describe an infrastructure-independent indoor localization approach for various indoor environments. Our method introduces a novel particle filter implementation that enables the fusion of inertial motion unit sensors, user context, user gait direction, and map information. Due to this novel fusion, it performs localization with up to two orders of magnitude fewer particles than state-of-the-art approaches. Additionally, it extracts map information via existing open services, such as the Open Street Maps and it follows defined standards for the map handling. We evaluated all the components of our method in realtime in off-the-shelf smartphones and we find that it performs a median error of 2.3m, while using only 40 particles instead of 400 or up to 4000 particles that other methods require for the same accuracy.

Crowdsourcing is a powerful technique for bootstrapping sensing systems that are based on wireless signals. For example, wireless sensing systems can ask users to contribute training data and localization systems (such as WiFi fingerprinting) can take advantage of wireless measurements provided by users of the system. Indeed, previous research has demonstrated that crowdsourcing can reduce labor efforts needed to deploy and initialize wireless sensing systems by several orders of magnitude, without compromising on system performance. Despite the many benefits of crowdsourcing, current methods suffer from one significant drawback, namely that they are highly sensitive to variations in devices capturing the measurements. Indeed, as we demonstrate in this paper, cross-device variations can decrease performance of crowdsourced bootstrapping approaches up to 70of radio maps used for localization and to make them robust against cross-device variations in wireless signals. We evaluate our framework by considering WiFi fingerprinting based localization as a representative example of applications that benefit from our approach. Our results demonstrate up to 1.8m localization error, and 18.7.

Semantics associates meaning with Internet of Things (IoT) data and facilitates the development of intelligent IoT applications and services. However, the big volume of the data generated by IoT devices and resource limitations of these devices have given rise to challenges for applying semantic technologies. In this article, we present Cloud and edge based IoT architectures for semantic reasoning. We report three experiments that demonstrate how edge computing can facilitate IoT systems in terms of data transfer and semantic reasoning. We also analyze how distributing reasoning tasks between the Cloud and edge devices affects system performance.

Consumption of data streams generated from IoT devices during IoT application development is gaining prominence as the data insights are paramount for building high-impact applications. IoT mashup tools, i.e. tools that aim to reduce the development effort in the context of IoT via graphical flow-based programming, suffer from various architectural limitations which prevent the usage of data analytics as part of the application logic. Moreover, the approach of flow-based programming is not conducive for stream processing. We introduce our new mashup tool aFlux based on actor system with concurrent and asynchronous execution semantics to overcome the prevalent architectural limitations and support in-built user-configurable stream processing capabilities. Furthermore, parametrizing the control points of stream processing in the tool enables non-experts to use various stream processing styles and deal with the subtle nuances of stream processing effortlessly. We validate the effectiveness of parametrization in a real-time traffic use case.

In this paper we present our developed and evaluated method for the dynamic mapping of the vertical characteristics inside a building. For achieving that, we extract data from smart-phone sensors and use those data for altitude estimation via the barometric formula. We introduce a novel approach for the extraction of reference pressure during the outdoor-to-indoor-transition of the user inside a building, which is identified through sensor fusion. A combination of machine learning techniques is used for the identification of the number of floors and the unsupervised classification of the altitude of each floor. As far as we know, this is the first system able of mapping vertical characteristics inside a building autonomously. Finally, enhancements on the CityGML model are made for mapping those characteristic by following its given standards.

We present a new method to translate the established and well-known ConcurTaskTree (CTT) task modeling technique into state machines. For this purpose, we develop the concepts of partial state machines, Connectables and a connect operator, which form the theoretical framework for a new algorithm. For the translation, we develop and present a recursive, bottom-up algorithm, which exploits the inherent structure of CTTs and produces valid, partial state machines. This allows new development processes in the model-driven application and system development domain.

This paper presents our research in developing a model for the dynamic generation of indoor maps with crowdsourcing. With approximation of the user traces, we generate a point cloud and develop the topology of the space from time based segmentation of the traces. Moreover, we add semantic information for navigation and localization enabled maps. We discuss motivation, research objectives, and detailed research methods in this paper.

The advent of Big Data has created a rich set of diverse languages and tools for data manipulation and analytics within the Hadoop ecosystem. Pig has a prominent role within this ecosystem as a scripting layer-a convenient way to create analytics jobs that are issued for batch processing in a Hadoop cluster. In order to leverage the benefits of graphical domain specific languages, namely intuitive visual design and inspection, we implemented a web-based graphical tool called QryGraph that complements Pig in various ways. First, it allows a user to create Pig queries in a graphical editor and check their syntax. Second, it provides an administrative interface for managing the execution and overall lifecycle of Pig queries. Finally, it will allow for debugging by running queries on test data sets and for creating user-defined query sub-graphs that can be reused across different Pig queries.

In order to cope with changing markets and reduced product life-cycles production systems need to get more adaptive and allow to easily introduce new products and product variants. Current solutions require that all control equipment need to be modified in such cases. In this work we propose a different approach with a strong separation of basic control of production resources from product specific behavior. The product specific behavior is managed by a central control entity which sends parts of the execution tree required for a certain to thin controllers managing the production resources. With this approach we can on one hand achieve the needed flexibility and at the other hand side achieve the timing requirements of production.

In this paper, we focus on networked, embedded systems which may contain numerous electronic control units, connected by multiple network busses. Furthermore, such embedded systems support many runtime configurations. The main problem is to determine the network resource needs of all variations permitted at runtime, i.e. to calculate the worst case resource needs. We describe the runtime variability of such systems by means of a runtime feature model and then derive a network performance model in a stepwise way. The mappings from the feature level leads to a data flow model, then to the component/hardware level, where we perform a detailed network analysis based on Network Calculus. Thus, we can decide in an early design stage, whether a given network topology fits the requirements of a given software architecture. We show then, by an example, (i) that the performance of this component/hardware model can be analyzed using network calculus and (ii) that our approach can significantly reduce resource overestimation compared to a static evaluation.

In this work, we consider reliable runtime adaptation in networked, embedded systems with tight real-time constraints. Specifically, we adapt the placement of software components on a multitude of hardware components and show the need for a hierarchical transaction concept in this context. We consider multiple adaptation steps under these hard system constraints and also introduce a model with undesired configurations, which cannot be maintained for an extended time period. Furthermore, we identify cases when adaptation steps can be performed in parallel. Our concept for adaptations is applied to realistic automotive examples, where the feasibility of adaptation in the inactive period of a software component is demonstrated.

In software engineering, design patterns are commonly used and represent robust solution templates to frequently occurring problems in software design and implementation. In this paper, we consider performance simulation for two design patterns for processing of parallel messaging. We develop continuous-time Markov chain models of two commonly used design patterns, Half-Sync/Half-Async and Leader/Followers, for their performance evaluation in multicore machines. We propose a unified modeling approach which contemplates a detailed description of the application-level logic and abstracts away from operating system calls and complex locking and networking application programming interfaces. By means of a validation study against implementations on a 16-core machine, we show that the models accurately predict peak throughputs and variation trends with increasing concurrency levels for a wide range of message processing workloads. We also discuss the limits of our models when memory-level internal contention is not captured.

Trust plays an important role in software systems, especially component-based systems in which components or their environments vary. This paper introduces an autonomic trust management solution for a component-based software system. We propose an adaptive trust control model to specify, evaluate, establish, and ensure the trust relationships among system entities. This model concerns the quality attributes of the entity and a number of trust control modes supported by the system. In particular, its parameters can be adaptively adjusted based on runtime trust assessment in order to reflect real system context and situation. Based on this model, we further develop a number of algorithms that can be adopted by a trust management framework for autonomic management of trust during component execution. We verify the algorithms' feasibility through simulations and demonstrate the effectiveness and benefits of our solution. We also discuss the issues for successful deployment of our solution in a component software platform.

Abstract This article compares the organization and practices for software reuse in integration‐oriented software product lines (SPLs) and open source software projects. The main observation is that both approaches are successful regarding large variability and reuse, but differ widely in their practices and organization. To capture practices in large open source projects, we describe an open compositional model, which reflects their more decentralized organization of software development. We capture key practices and organizational forms for this and validate these by comparing four case studies of this model. Two of these studies are based on published SPL case studies, for the other two we analyze the practices in two large and successful open source projects based on their published developer documentation. Our analysis highlights key differences between the practices in the two open source organizations and the more integrational practices used in the other two cases. Finally, we discuss which practices are successful in which environment and how the current practices can move towards more open, widely scoped and distributed software development. Copyright © 2010 John Wiley & Sons, Ltd.

This paper presents our approach for building an internet based middleware platform for smart spaces as well as a number of services and applications that we have developed on top of it. We outline the architecture for the smart space middleware and discuss several applications and services that we have so far realized with this middleware. The presented material highlights key concepts in our middleware vision: services are HTTP based and restful; applications are accessed through a browser so that they are available on a wide variety of devices; and we demonstrate the concept of bridging non internet enabled smart space devices to our IP and HTTP centric smart space network.

In this paper, we consider semantic refinement for feature-oriented programming where components are built from features and weavings, which we use to adapt one feature to the context of another one. We address the question of semantic reasoning about multiple weavings. If we know the effect of feature A on X and of feature B on X, what can we conclude about adding both A and B to X? For this, we define conservative weavings which do not modify the state of another feature. We show that composition of several such weavings is however not compositional as it does not preserve semantics. In particular, weavings must consider that other weavings have already been applied. This explains why it is considerably more difficult to reason about multiple aspect weavings. We show criteria on the dependencies between weavings which allow modular, semantics-preserving application of weavings. This is formalized in a calculus for feature composition and also extended to conditional refinements.

When an ad hoc network is connected to the Internet, it is important for the mobile nodes to detect available gateways providing access to the Internet. We discuss the problem of interworking and integration between the ad hoc and radio access networks. Furthermore, we propose a signalling mechanism to discover and select the optimum gateway when more than one gateway is available. A gateway is considered to be optimum if the path via this gateway provides better quality of service such as total path bandwidth or overall delay. An optimum gateway selection scheme should consider both the ad hoc and radio access networks to provide better QoS support. To achieve this selection, QoS information from both sides of the network needs to be communicated across the network. Further research is required to adapt the proposed mechanism to be applied for wireless sensor networks

This paper presents an end-to-end reservation protocol for quality-of-service (QoS) support in the medium access control layer of wireless multihop mesh networks. It reserves periodically repeating time slots for QoS-demanding applications, while retaining the distributed coordination function (DCF) for best effort applications. The key features of the new protocol, called "distributed end-to-end allocation of time slots for real-time traffic (DARE), are distributed setup, interference protection, and scheduling of real-time data packets, as well as the repair of broken reservations and the release of unused reservations. A simulation-based performance study compares the delay and throughput of DARE with those of DCF and the priority-based enhanced distributed channel access (EDCA) used in IEEE 802.11e. In contrast to DCF and EDCA, DARE has a low, nonvarying delay and a constant throughput for each reserved flow

Efficient support for multimedia communication is essential for next generation mobile wireless networks. In order to provide the necessary quality-of-service (QoS) in mobile wireless networks, appropriate signaling has to be defined. Our QoS signaling architecture integrates resource management with mobility management. It is based on a domain resource manager concept and nicely supports different handover types. In particular, our approach supports anticipated handover with pre-reservation of resources before the mobile node is attached to the new access point. The article discusses the protocol design and implementation, in particular integration with mobile IP, and shows the benefit of anticipated handover by measurement results.

When an ad hoc network is connected to the Internet, it is important for the mobile nodes to detect the available gateways providing access to the Internet. Therefore, a gateway discovery mechanism is required. The two main approaches for discovering Internet gateways are the reactive and the proactive one. This paper compares the performance of these approaches in various scenarios by means of simulation. We show that the proactive approach performs better in the simulated scenario.

This paper addresses the connectivity between ad hoc networks and IP based fixed networks. This connection can be achieved through an Internet gateway, which in this paper is assumed to be a router at the edge of the fixed network. We investigate issues on Internet gateway discovery, addressing and QoS support for interconnecting ad hoc networks and the Internet. We compare different approaches regarding addressing and micro-mobility management. Our simulation results show the performance of an ad hoc routing protocol while connecting to the Internet.

Traditionally, handover decisions are made mainly based on the signal strength. However, with the evolution of mobile networks moving toward a heterogeneous solution, this simple decision mechanism is not enough any more. To choose the optimal access router among the multitude points of attachment available to the network, there is a need for a more intelligent handover decision mechanism. In this paper we provide a general framework for handover decisions which take into account the context of both the mobile network and the user. Our solution shows how to collect the context, to compile information, and to propagate it proactively. Due to the fast evolving nature of context information, our flexible framework allows one to update the decision algorithm and the context data. We present the architecture and the main components of our proposed implementation based on active networks.

Abstract We propose a new model for flexible composition of objects from a set of features. Features are services of an object and are similar to classes in object‐oriented languages. In many cases, features have to be adapted in the presence of other features, which is also called the feature interaction problem. We introduce explicit interaction handlers which can adapt features to other features by overriding methods. When features are composed, the appropriate interaction handling is added in a way which generalizes inheritance and aggregation. For a set of features, an exponential number of different feature combinations is possible, based on a quadratic number of interaction resolutions. We present the feature model as an extension of Java and give two translations to Java, one via inheritance and the other via aggregation. We show that the feature model interacts nicely with several common language extensions such as type parameters, exceptions, and higher‐order functions. Copyright © 2001 John Wiley & Sons, Ltd.

Functional logic languages with a sound and complete operational semantics are mainly based on an inference rule called narrowing. Narrowing extends functional evaluation by goal solving capabilities, as in logic programming. Due to the huge search space of simple narrowing, steadily improved narrowing strategies have been developed in the past. Needed narrowing is currently the best narrowing strategy for first-order functional logic programs due to its optimality properties wrt the length of derivations and the number of computed solutions. In this paper, we extend the needed narrowing strategy to higher-order functions and λ-terms as data structures. By the use of definitional trees, our strategy computes only independent solutions. Thus, it is the first calculus for higher-order functional logic programming which provides for such an optimality result. Since we allow higher-order logical variables denoting λ-terms, applications go beyond current functional and logic programming languages. We show soundness and completeness of our strategy with respect to LNT reductions, a particular form of higher-order reductions defined via definitional trees. A general completeness result is only provided for terminating rewrite systems due to the lack of an overall theory of higher-order reduction which is outside the scope of this paper.

Abstract We study the type inference problem for a system with type classes as in the functional programming language Haskell. Type classes are an extension of ML-style polymorphism with overloading. We generalize Milner's work on polymorphism by introducing a separate context constraining the type variables in a typing judgement. This leads to simple type inference systems and algorithms which closely resemble those for ML. In particular, we present a new unification algorithm which is an extension of syntactic unification with constraint solving. The existence of principal types follows from an analysis of this unification algorithm.

Introduces several approaches for solving higher-order equational problems by higher-order narrowing, and gives some completeness results. The results apply to higher-order functional-logic programming languages and to higher-order unification modulo a higher-order equational theory. We lift the general notion of first-order narrowing to so-called higher-order patterns and argue that the full higher-order case is problematic. Integrating narrowing into unification (called 'lazy narrowing') can avoid these problems and can be adapted to the full higher-order case. For the second-order case, we develop a version where the needed second-order unification remains decidable. Finally, we discuss a method that combines both approaches by using narrowing on higher-order patterns with full higher-order constraints.>

We study the type inference problem for a system with type classes as in the functional programming language Haskell. Type classes are an extension of ML-style polymorphism with overloading. We generalize Milner's work on polymorphism by introducing a separate context constraining the type variables in a typing judgement. This lead to simple type inference systems and algorithms which closely resemble those for ML. In particular we present a new unification algorithm which is an extension of syntactic unification with constraint solving. The existence of principal types follows from an analysis of this unification algorithm.