Marisol García-Valls

Challenges in real-time virtualization and predictable cloud computing

Cloud computing and virtualization technology have revolutionized general-purpose computing applications in the past decade. The cloud paradigm offers advantages through reduction of operation costs, server consolidation, flexible system configuration and elastic resource provisioning. However, despite the success of cloud computing for general-purpose computing, existing cloud computing and virtualization technology face tremendous challenges in supporting emerging soft real-time applications such as online video streaming, cloud-based gaming, and telecommunication management. These applications demand real-time performance in open, shared and virtualized computing environments. This paper identifies the technical challenges in supporting real-time applications in the cloud, surveys recent advancement in real-time virtualization and cloud computing technology, and offers research directions to enable cloud-based real-time applications in the future.

Guidelines for a graduate curriculum on embedded software and systems

The design of embedded real-time systems requires skills from multiple specific disciplines, including, but not limited to, control, computer science, and electronics. This often involves experts from differing backgrounds, who do not recognize that they address similar, if not identical, issues from complementary angles. Design methodologies are lacking in rigor and discipline so that demonstrating correctness of an embedded design, if at all possible, is a very expensive proposition that may delay significantly the introduction of a critical product. While the economic importance of embedded systems is widely acknowledged, academia has not paid enough attention to the education of a community of high-quality embedded system designers, an obvious difficulty being the need of interdisciplinarity in a period where specialization has been the target of most education systems. This paper presents the reflections that took place in the European Network of Excellence Artist leading us to propose principles and structured contents for building curricula on embedded software and systems.

QoS-Aware Real-Time Composition Algorithms for Service-Based Applications

This paper presents a model for quality-of-service (QoS)-aware service composition in distributed systems with real-time and fault-tolerance requirements. This model can be applied in application domains like, for example, remote monitoring, control and surveillance. Classic approaches to real-time systems do not provide the flexibility and fault-tolerance required in new emerging environments that need to combine a high degree of dynamism with temporal predictability. Our approach addresses these new challenges by combining concepts from the service oriented paradigm and distributed real-time systems. We propose a concrete system model based on a holistic time-triggered-based approach for design and configuration. Based on this model, we propose two algorithms for the composition of QoS-aware service-based applications with temporal requirements: an exhaustive algorithm that computes the optimal service combination in terms of a figure of merit, suitable for offline composition; and an improved algorithm based on heuristics and partial figures of merit, suitable for online composition. Experimental results show that the latter reduces dramatically the number of combinations explored with a minimal degradation in the quality of the solution, making it feasible for online execution in dynamic environments.

Real-time reconfiguration in multimedia embedded systems

This paper presents a software framework containing a reconfiguration mechanism that provides realtime execution guarantees in intensive media processing consumer electronic devices. Our approach offers real-time guarantees during all of the system operation, i.e., not only during the normal media processing mode but also in the event of reconfigurations (a reconfiguration is a transition among two different system states). This approach enhances the budget scheduling model to provide a time-partition that guarantees the availability of resources during transitions. Results are presented that validate the proposed approach.

Simple Asynchronous Remote Invocations for Distributed Real-Time Java

More and more, the use of real-time distribution middlewares programmed with high-level languages like Java is becoming of interest for industrial systems because this type of infrastructures reduce development efforts required to both design and maintain complex networked applications. In that way towards having better development tools, this paper deals with the very specific issue of the asynchronism from the particular perspective of Javas distributed technologies. It proposes and evaluates an extension to the Javas remote method invocation (RMI) with additional support for asynchronous remote invocations. Such a mechanism makes the transmission of some messages, typically those that do not require any confirmation from the server, much more efficient than synchronous communications, and consequently they optimize and speed up the performance of many industrial networked applications. In more detail, this paper concerns with the model itself and the set of changes it requires within their middleware programming interfaces and inside communication protocols to accommodate the proposed model. The proposal is supported by empirical evidences from a prototype developed as partial proof-of-concept.

Scheduling component replacement for timely execution in dynamic systems

Timely run‐time software replacement techniques are a corner stone for reconciling real‐time systems development and dynamic behavior. Typical real‐time systems do not consider dynamic behavior because it deeply challenges predictability and timeliness. Current efforts are starting to merge the safe and predictable execution with a controllable level of dynamicity by imposing a set of bounds and limitations to the system dynamic behavior. One of the obstacles for this is how to time‐bound the different operations required to effectively implement component replacement. In this paper, the main challenges for this problem are identified, and a model to ensure that components can be replaced at run time preserving the temporal properties of the system is provided that also avoids failures in replacements. A real example and simulations of our replacement model are provided that validate the presented ideas. Copyright

Towards the integration of scoped memory in distributed real-time Java

This paper presents a memory management solution for distributed real-time Java. The proposed model is targeted at distributed applications that require to keep their internal state and that consume a bounded memory size in their operation. The aim is to provide an alternative to the garbage collector for distributed real-time systems and to avoid the penalties it introduces. The approach is based on the usage of scoped memory defined in the real-time specification for Java (RTSJ). Our solution introduces constraints in the programming model and in the current distributed Java architecture. However, these are compensated by eliminating the dependency of the garbage collector in the end-to-end invocation and by achieving self-cleaning with respect to the memory consumed by remote invocations.

A Synchronous Scheduling Service for Distributed Real-Time Java

Current trends in real-time systems identify Java as a new alternative to develop both centralized and distributed real-time systems. Many efforts have been devoted to develop the Real-Time Specification for Java (RTSJ), and there is substantial ongoing activity to produce a straightforward and valuable Distributed Real-Time Specification for Java (DRTSJ). The current paper provides a contribution to this latter activity defining, from different angles, a synchronous scheduling service aligned with principles of some popular real-time architectures. This service orchestrates the system in such a way that it provides end-to-end guarantees in the distributed transactions, guaranteeing their timely execution across the network and nodes. The service is described from two points of view: the system one, characterizing a portable model; and the programmer one, defining a distributed object-oriented implementation of a model based on Real-Time Remote Method Invocation (RTRMI). Finally, it also presents results of an implementation carried out to judge the efficiency of the service, offering a preliminary predictability and performance assessment of a distributed real-time Java technology.

Towards a Synchronous Scheduling Service on Top of a Unicast Distributed Real-Time Java

This paper describes an approach towards the definition and implementation of a synchronization service on top of a remote object model offered by an unicast real-time remote object paradigm. Also an architecture model based on the RTSJ (real-time specification for Java) and the distribution middleware RMI (remote method invocation) specifications is proposed in order to give support to the model, defining a convergence layer that manages the underlying resources involved in a master-slave communication through a new API. Finally, preliminary results from an implementation prototype show the feasibility of the model and provide an initial estimation of jitters and the performance of the synchronization service

Adaptive real-time video transmission over DDS

An increasing number of industrial applications include video processing capacities, which allow, among others, remote monitoring of industrial processes and control of private and public areas. Image processing has real-time requirements which result in resource demands at both node and network levels. Moreover, video is usually compressed and coded to be transmitted which generates variable bit-rate streams. This introduces variable processing requirements inside the node in terms of memory and processor cycles required for the processing of the sequence of different video frames. A direct impact on the network resource is also obvious since variable network bandwidth will be required to transmit the frames that may affect the bandwidth assigned to other streams. Efficient distributed video surveillance requires that real-time constraints be respected or, at least, quality of service guarantees (QoS) be provided. The traditional approach to video transmission has focused at the level of the network protocols. However, architectural solutions at the middleware level introduce higher flexibility and more efficiency in development time. This paper presents an architecture that precisely defines an integral set up of the different components that are relevant in achieving real-time and QoS-based video surveillance. The paper describes how the DDS standard for real-time distributed systems, can be used for this purpose; based on the decoupled interaction paradigm of DDS, higher complexity surveillance deployments are possible. A prototype surveillance system is presented which includes video transmission and adaptation to environmental sensing events.