The graphical modelling languages UML and SysML, nowadays widely used in industry, integrate different modelling concepts and notations in one standardised framework. However, they lack a clearly defined, unambiguous semantics and thus their formal verification represents a challenge. On the other hand, current safety standards, including ISO 26262, demand such verification especially for safety-relevant systems. The literature proposes a plethora of different semantics and formalisms for UML/SysML. In this paper we compare and summarise existing work on the formalisation of behavioural UML and SysML models and their verification. Our goal is to foster a better understanding of the problems related to UML/SysML formalisation, and to aid people bridging the gap from high level graphical modelling to formal verification techniques.

Smart connected electromobility will leverage the cross-domain cooperation along a new value chain of stakeholders from the automotive and energy industry. Intelligent charge management, i.e. charging the electric vehicle as a tradeoff between charging costs and personal or operational constraints, represents an important service in an electronic service market for electromobility. The market features the energy supplier as a service provider, the electric vehicle as a service consumer along with intermediate brokers in between. All of them continuously interact to exchange data, thus implementing the vision of an open and global Internet of Services in an electromobile ecosystem. A joint service life cycle management which considers both, the build-time and the runtime view of a service, is essential for such a service market. It allows the service provider to design, implement and deploy services while investigating systems of deployed services and their on-demand consumption at run-time. In this paper we present such an integrated service life cycle model for electronic electromobility market places. The life cycle model will be elaborated by means of an intelligent charge management use case.

Reconfigurable hardware is gaining a steadily growing interest in the domain of space applications. The ability to reconfigure the information processing infrastructure at runtime together with the high computational power of today's FPGA architectures at relatively low power makes these devices interesting candidates for data processing in space applications. Partial dynamic reconfiguration of FPGAs enables maximum flexibility and can be utilized for performance increase, for improving energy efficiency, and for enhanced fault tolerance. To be able to prove the effectiveness of these novel approaches for satellite payload processing, a highly scalable prototyping environment has been developed, combining dynamically reconfigurable FPGAs with the required interfaces such as SpaceWire, MIL-STD-1553B, and SpaceFibre. Up to 30 SpaceWire interfaces, 5 copper-based SpaceFibre interfaces, and 270 GPIOs can be realized and combined with one to five dynamically reconfigurable Xilinx FPGAs and up to 20 GByte of working memory. The implemented approach for dynamic reconfiguration enables partial reconfiguration at 400 MByte/s. Blind and readback scrubbing is supported and the scrub rate can be adapted individually for different parts of the design.

In the digital product development of the automotive and aviation industries, the traditional digital mock-up (DMU) is about to be supplemented and superseded by the digital functional mock-up (DFMU). The DFMU extends the geometric integration of product models, as provided by traditional DMUs, with functional aspects. Thus, it allows investigating both physical and logic model properties. Additionally, DFMUs build the foundation for an integrated analysis of holistic behavioural aspects of a product. The design of DFMUs is validated by performing a cost-efficient co-simulation — a methodology for simulating individual components concurrently by different simulation tools which mutually exchange information in a collaborative manner. This process can be highly cross-organisational and thus requires an exchange of cost-sensitive and confidential product models between suppliers and OEMs. This paper discusses the deployment of service-oriented architectures in co-simulation environments for (a) improving the configuration process, (b) protecting the intellectual property of the model owners, and (c) introducing electronic signatures to simulation models, data and results. Furthermore, it presents the concept of a functional mock-up trust centre (FMTC) which implements part of these services. The FMTC is a practical and effective instrument to preserve intellectual property rights of model providers during co-simulation. This is achieved by encrypted storage facilities for product models and a safeguarded simulation environment.

Sensor technology continues to improve at the price of increased data rates, which require being processed. In the space domain, the available bandwidth for effectively transferring the data to the base station is limited, such that there is a need for a high-performance data processing unit on board of the spacecraft. This work targets the development of a scalable high-performance payload data processing system based on dynamically reconfigurable FPGAs. The system, which is called Dynamically Reconfigurable Processing Module (DRPM), enables a multitude of high performance data processing applications to be supported by the same hardware in space. While reconfigurable hardware offers higher performances than traditional DSP-based solutions, it also supports the same flexibility to modify the functionality at run-time. Dynamically Reconfigurable Processing Module (DRPM) The DRPM is a multi-FPGA architecture, which is designed especially for space applications. The FPGAs (Xilinx Virtex-4 family) are used to implement high performance data processing cores for a wide range of applications. In addition, the DRPM supports in-flight reconfiguration during a mission where required, whilst being fault-tolerant to space environment effects typically caused by high energy particles. Figure 1 shows a block diagram of the DRPM. The DRPM can be divided into interface components, control components, and data processing components. The payload data processing is performed on the FPGAs. Each FPGA is connected to a local memory bank, which is used for storing application data. Most of the FPGA resources are used for reconfigurable modules, which perform the payload data processing. Besides the reconfigurable modules each FPGA implements a memory controller for the local memory and a dynamic processing control unit to manage addresses and resources of the reconfigurable modules. The DRPM supports a variety of interfaces (CAN, SpaceWire, MIL-STD1553, etc.) to establish connection to avionics and source data instruments. The system controller is based on a Leon2-FT processor and offers services to the ground station to control all functions executed on the DRPM. The reconfiguration controller performs the reconfiguration processes of the FPGAs. Reconfiguration is used to M. Koester, J. Hagemeyer, F. Margaglia and M. Porrmann are with Heinz Nixdorf Institute, University of Paderborn, Paderborn, Germany. F. Dittmann and M. Ditze are with TWT GmbH Science & Innovation, Neuhausen, Germany. L. Sterpone is with Politecnico di Torino, Torino, Italy. J. Harris is with Swiss Space Technology, Champery, Switzerland. J. Ilstad is with ESTEC, Noordwijk, The Netherlands. Control Components Data Processing Interfaces

This paper presents an adaptive near-optimal scheduler for multimedia traffic for the 802.11e Enhanced Distributed Channel Access (EDCA) medium access control scheme: The scheduler exploits the ant colony optimization (ACO) meta heuristic to tackle the challenge of packet scheduling. ACO is a biologically inspired algorithm that is known to find near-optimal solutions for combinatorial optimization problems. Thus, we expect that ACO scheduling produces more efficient schedules than comparable deterministic scheduling approaches at the expenses of a computational overhead it introduces. We compare ACO scheduling relevant deterministic scheduling approaches, and in particular the MLLF scheduler that is specifically designed for the needs of compressed multimedia applications. The purpose of the evaluation is twofold. It allows to draw conclusions on the feasibility of ACO scheduling for multimedia traffic while it serves as a benchmark to determine to what extent deterministic schedulers fall short of a near-optimal solution.

Admission Control (AC) avoids network congestion by determining whether a quantified request for resources can be approved without interfering the resource allocation of already accepted traffic flows. For time sensitive networked multimedia applications conventional AC schemes often prove to be unnecessarily strict. The crisp binary admission decision and worst case assumptions on traffic behavior may result in low utilization of resources. In this paper we propose a less stringent fuzzy logic based solution for AC. It decides an admission request according to a collective set of fuzzified Quality of Service (QoS) parameters that anticipate the expected resource utilization. Furthermore, it maps qualitative and quantitative QoS parameters into a single decision metrics. The AC has been incorporated into the service oriented UPnP QoS architecture.

Real-time applications usually are executed on top of a Real-time Operating System (RTOS). Specific scheduling algorithms can be designed. When possible, static cyclic schedules are calculated off-line. If more flexibility is needed on-line techniques are applied. These algorithms are bound to priorities which can be assigned statically or dynamically. Designing a proper RTOS architecture needs some delicate decisions. The basic services like process management, inter-process communication, interrupt handling, or process synchronization have to be provided in an efficient manner making use of a very restricted resource budget. Various techniques like library-based approaches, monolithic kernels, microkernels, or virtual machines/exokernels are applied, based on specific demands. Safety critical application can be supported by separation of applications either in the time or the space domain. Multi-core architectures need special techniques for process management, memory management, and synchronization. The upcoming Wireless Sensor Networks (WSN) generate special demands for RTOS support leading to dedicated solutions. Another special area is given by multimedia applications. Very high data rates have to be supported under (soft) RT constraints. Based on the used encoding techniques (e.g. MPEG) dedicated solutions can be created.

The transmission of time-sensitive multimedia traffic across unreliable wireless network links requires adaptive scheduling that respects substantial resource fluctuations and irregular traffic arrival patterns. As finding a near-optimal adaptive transmission schedule, in our case for the prioritized access categories of the 802.11e Enhanced Distributed Channel Access (EDCA), is a combinatorial optimization problem, we introduce an Ant Colony Optimization (ACO) based meta heuristic for scheduling multimedia traffic. ACO is known to develop near-optimal solutions at the expenses of additional administrative and computational overhead. We compare ACO scheduling to a deterministic scheduling approach specifically designed for the needs of multimedia scheduling called MLLF. The aim of this paper is to evaluate the feasibility of ACO approaches for multimedia scheduling in different scenarios and to conclude on the effectiveness of the selected deterministic MLLF approach in comparison to a near-optimal solution.

Distributed multimedia applications increasingly populate Audio-Visual (AV) mobile and desktop devices. Crucial to the success of these applications is the delivery of Quality of Service (QoS) which often refers to resource reservation on devices and network links. Resource reservation is usually ensured by an appropriate admission control (AC) that determines if another request for resources can be granted without interfering already accepted traffic flows and processing resources. If an AC cannot be processed or cannot grant the requested resources, proactive measures like resource adaptation may control the resource requirements of particular flows. Resource adaptation adjusts the resource demands of the application according to the available resources. This paper presents a flexible and hybrid resource adaptation framework for multimedia applications that is incorporated into the UPnP QoS architecture. Furthermore, a simple methodology for controlling resource adaptation in mobile networks and devices is presented. An implementation of resource adaptation proves that the resource utilization is optimized and the QoS can be effectively maintained even in overloaded conditions.

Distributed multimedia applications increasingly populate audio-visual mobile and desktop devices. Crucial to the success of is the delivery of quality of service (QoS) which often refers to resource reservation on devices and network links. QoS is usually ensured by an appropriate admission control that determines if another request for resources can be granted without interfering already accepted data flows and processing resources. If an AC cannot be processed or proves to be unsatisfactory, proactive counter measures like resource adaptation may decrease the resource requirements of particular data flows. This paper presents a new methodology for resource adaptation in AV devices. It is incorporated into the UPnP QoS architecture and concealed behind a common set of interfaces. An implementation of resource adaptation proves that the resource utilization is optimized and the QoS can be acceptably maintained even in overloaded conditions.

Multimedia Applications e.g. Video/Audio streaming increasingly strive for deployment in distributed and domain spanning networking environments. In order to ensure the service interoperability between content servers and rendering devices across heterogeneous platforms, distributed middleware approaches offer promising concepts (e.g. Service Oriented Architectures). The UPnP forum has developed a platform-neutral architectural framework that specifies a set of well defined service discovery and access routines concealed behind a unified set of protocols and their APIs. While UPnP supports service interoperability additional service extensions are required to satisfy requirements as imposed by many multimedia applications. These requirements cover QoS support from the application source to the sink which entails the encapsulation of signalling services that allow for resource reservation on network and end-devices. Furthermore we introduce a proactive content replication paradigm: According to popularity forecasts it is possible to create replicas of e.g. movie content to avoid hot-spots maintaining very popular content in large scale networks. The contribution of this paper is to identify domain spanning distributed multimedia scenarios and describe the challenges for developing enhancing services to the UPnP framework that help to establish them.

The upcoming IEEE 802.11e standard improves the Medium Access Control (MAC) of the legacy 802.11 with regard to Quality of Service (QoS) by introducing the Enhanced Distributed Channel Access (EDCA) and the HCF Controlled Channel Access (HCCA). EDCA achieves QoS by providing independent transmit queues and MAC parameters for each traffic class, and hence higher prioritized traffic has a higher probability for transmission. Crucial to the success of such a strategy is a scheduler that assigns the data traffic to the respective transmit queues. This paper develops and accommodates a new dynamic scheduler for EDCA into the MPEG-4 Delivery Framework. Experiments prove that the new scheduling policy timely delivers up to 50% more frames than statical scheduling solutions. To the best of our knowledge this is one of very few scheduling approaches that considers MPEG-4 related traffic priorization in EDCA.

Video-on-demand applications and their corresponding issues have been a long-term research topic. The commercial deployment and exploitation of these scenarios in PAEANs (public authorized enterprise access networks) like in e.g. hotels, schools or airport lounges, however, is still in its early stages. The reason for that are the high associated costs, the huge amount of bandwidth requirements for high-quality video combined with the low penetration of broadband access and the leaking quality of service (QoS) support. We proposed peer-to-peer technology that overcomes these drawbacks. Our approach allows each peer in the network to act as a restricted video-server that offers a low number of selected videos to other peers. This decentralized approach overcomes the traditional drawbacks of client-server based solutions. Crucial to the success of this approach, however, is an intelligent content replication strategy that ensures that highly frequented content is replicated across the network. This allows for innovative workload scaling to avoid local bottlenecks. Furthermore, our architecture allows to transmit this content in quasi real-time. Hence, we present a new QoS-oriented architecture that works in concert with the above mentioned concepts. It uses true point-to-point communication in order to preserve the on-demand character of the applications.

Video streaming applications (e.g. video conferencing, video-on-demand) increasingly strive for deployment in small embedded systems that traditionally exhibit small computational resources as well as low speed internal network connections, e.g. set top boxes, mobile phones or PDAs. Whereas the latter is addressed by the new scalability features in the MPEG-2 and MPEG-4 standards, the computational resources still must be used effectively, all the more as continuously improved compression algorithms increase the computational demands. In order to guarantee a frame-per-second rate that satisfies the requested Quality of Service (QoS), a real-time scheduling mechanism is required that accounts for the specific needs of the respective compression standards along with a corresponding mechanism for Admission Control. The introduction of different frame-types in MPEG that require varying resources for decoding and the possibility of variable-bit-rate encoding, however, result in strong workload imbalances and unpredictability that do not allow to exploit the available resources efficiently. Therefore, this paper presents a new approach that allows to balance the workload caused by MPEG stream decoding. It allows to timely decode an additional 16% of frames compared to traditional solutions. Furthermore, we introduce a method called Peak Notification that may reduce resource over-reservation by considerable 67% through workload peak prediction compared to common solutions. Both methods increase the QoS delivered to the client. To our knowledge, this is the first approach that balances the workload on a single processor to achieve better CPU utilization.

Peer-to-Peer networks have gained a lot of popularity in the past few years. Each participant offers own resources and occupies resources of other peers. This paper proposes an architecture for a P2P streaming network that considers structured data placement strategies of replicated multimedia content and QoS routing (RSVP). We consider point to point streaming within an autonomous system. So called active Rendezvous Servers do not just act as lookup services but rather perform the content distribution within the network. By analyzing access frequency time series it is possible to perform an intelligent replica method. Thus replica creation depends on access frequency, storage location, and each peers' admission control.