Today's system's requirements include distributed and net-centric concepts and capabilities. The challenge is to rapidly and accurately distribute large amounts of information to large numbers of nodes, over a variety of transports, account for application level Quality of Service (QoS) all while realizing a decoupling data-centric software architecture. Traditional point-to-point integration technologies and techniques simply don't work in these large distributed environments.

This tutorial introduces the OMG Data Distribution Service (DDS) and highlights the standard's unique capabilities that address and solve these real-time distributed system integration challenges. Specifically, the talk will include an introduction to publish/subscribe concepts, an overview of the specification, DDS data-modeling capabilities, and many of the application level QoS polices and their uses in real-world applications. The tutorial will conclude with a simple demonstration highlighting key concepts covered.
 

UPDM is the Unified Profile for DoDAF and MODAF. Based on UML and SysML, its designed to improve the integration between architectural framework modeling and system modelling to support post acquisition life-cycle design and implementation of complex systems. This tutorial provides a brief introduction to UPDM, an overview of the UPDM views and viewpoints and language concepts, and selected sample problems to demonstrate how the language can be used. We will demonstrate how the Model-Based Software Engineering foundation of UPDM provides answers to common real-world problems, such as avoiding the problems of stovepipe development, integrating requirements management into modelling, reusing architectures, communicating with non-experts and supporting the specification of real-time aspects at the Enterprise and Systems level.
 

The DDS specification provides fine-grained control over the real-time behaviour, dependability, and performance of DDS applications by means of a rich set of QoS Policies. The challenge for many DDS users is that the specifications explains very clearly how each QoS allows to control very specific aspects of data distribution yet it provides no hints on how different QoS should be composed to control complex properties such as the consistency model, or to impose end-to-end real-time scheduling decision. This half-day tutorial will fill this gap by providing attendees with (1) an explanation of how the various QoS compose, and (2) providing attendees with a series of QoS-composition Patters that can be used to control macro-properties of an application, such as the consistency model.
 

Supporting uninterrupted services for applications operating in resource-constrained distributed real-time and embedded (DRE) systems is hard due to the conflicting demands imposed by the timeliness and high availability quality of service (QoS) criteria on the available resources. Fault tolerance based on passive replication is deemed suitable for resource-constrained systems. Supporting real-time fault-tolerance using passive replication in highly dynamic DRE systems, where workloads and resource availabilities fluctuate constantly, requires sophisticated load-aware runtime mechanisms and algorithms. These mechanisms can impose excessive resource overhead, however, when operating in highly resource-constrained and closed systems with invariant system workloads, which calls for real-time and fault-tolerance solutions that can be conceived offline.

To address this dilemma (while supporting the needs of large-scale, DRE systems where network-resources must also be effectively allocated) this tutorial will describe a holistic, middleware-based solution comprising three parts: a design-time approach to resource allocation for DRE systems which a replica-to-node mapping algorithm, a model-driven network resource provisioning engine for real-time middleware such as RTCORBA, and an adaptive, load-aware middleware solution for real-time, fault-tolerance in open DRE systems that can easily be implemented for OMG's RTCORBA and CCM, and traditional CORBA facilities like Portable Interceptors.
 

TRACK 1

TRACK 1

This birds-of-a-feather session with be devoted to exploring the current state of the DDS technology. Some of the recent and upcoming standards as well as planned and desired future directions. Active community participation is expected with members of the community sharing some of their experiences and best ideas for future directions.
 

Safety evidence demonstrating the transformation process of Eurocat to be non-distortive of original functionality, convinced Thales customers to accept Eurocat modernization by means of a metrics-guided architecture driven modernization process and technology that achieved 100% automated code transformation accompanied by iterative semi-automated re-engineering to adapt Eurocat to meet the rigorous and exacting performance and architecture requirements of the next generation EATMS.

A twofold improvement in code and design quality metrics as measured by key design metrics was achieved during the transformation of Eurocat system software from mission-critical Ada 83 into high-performance, real-time Java. Today, the Eurocat system has been officially accepted by major ATMS customers, and will commence operation at the beginning of 2011 at ATC regions all across Europe and Asia. The transformed software has since then successfully passed several factory acceptance tests and has been deployed on sites. Manual refactoring operations, done to enable a standardised product management of fully independent modules has also be performed since the original transformation
 

The results presented, specifically the application monitoring meta-model and instrumentation API, could influence a future OMG monitoring and instrumentation standard. This elusive goal would greatly benefit distributed application users and is of special interest to the C4I community.
 

This goal of the presented effort is to evaluate and simplify the use of Data Distribution Service for astronomers not familiar with the DDS details. The evaluation includes various criteria: performance, ability to support the data structures common for astronomical data and ease of use. To facilitate the use of DDS, we developed wrapper classes around C++ bindings (generated by OpenSplice) and Python scripts, that minimize the familiarity with C++ mappings and expose a very simple C++ API. This API is also reflected in Python - a common language of choice for the astronomical applications.
 

Next Generation Network Services (such as VoIP, IPTV and, presence-based applications) and the so called Rich Communication Suite are driving the specification of the all-IP IMS architecture in which the main goal is to provide more than just voice for communication in fixed and mobile access networks. In this framework, the Session Initiation Protocol (SIP) is becoming the de-facto standard for session (call-control) signaling. In this presentation we propose the design of a DDS-SIP Gateway. This gateway will achieve interoperability of DDS data-spaces and SIP devices. We will describe the benefits of developing such a gateway, the adopted system design with comparison of different alternatives, and our recommendations for the evolution and extension of the DDS standards family in the light of our work.
 

SimD is an Open Source project that implements a simplified DDS API that allows a DDS application to be written with as little as 3 lines of DDS-specific code. This presentation will explain the rationale behind the design of SimD, cover the "object reference" approach used to represent all DDS entities, address the automatic memory management provided by SimD and some advanced techniques such as C++ type traits and templates to obtain automatic DDS types registration. In addition the presentation will cover in detail the new kinds of waitsets and listeners introduced by SimD to provide very efficient and type-safe waitsets and listeners.
 

This paper proposes a new co-simulation approach based on a solid OMG standard, the SysML State Machine Diagrams (for discrete events modeling) as implemented in Artisan Studio, and on the industry de-facto standard simulation tool Matlab Simulink (normally used for continuous time modeling). The main feature is based on an MBE approach: the automatic generation of optimized code, allowing real-time simulation that can run natively on a target for embedded systems. Additional features are present that enhance the effectiveness of the simulation, like remote graphical animation and control of the evolution of the state diagrams.
 

Reusable embedded components are the main subjects of component-based software development in embedded systems. Validating the functionality of an embedded component in component development phase by application developers increases the reusability of the component. For the components represented in UML, black-box (functional) testing of a component according to its UML model enables discovering functional faults and defects much earlier in the development lifecycle and makes testing easier by use of a common language with application developers. The UML Testing Profile provides concepts to develop test specifications and test models for testing UML model of a software system. In this presentation, a lightweight automated test framework based on the UML Testing Profile is proposed for efficient black-box testing of embedded components.
 

Application domains such as Airborne Networking (AN) or Intelligent Transportation Systems (ITS) require timely and reliable dissemination of information (e.g., timely delivery of traffic-related information to drivers or target information to pilots). Numerous impediments stemming from physics- and cyber-imposed factors lead to unpredictability in the timely and reliable dissemination of information. This presentation will illustrate these impediments to the timely and reliable dissemination of information based on results of simulations conducted in the OMNeT++/INETMANET simulator framework for IEEE 802.11 networks. A CPS solution to overcome these impediments and realized in the context of intelligent transportation systems will be presented. In particular, the CPS solution involves algorithms that derive insights from the experimental evaluations. Finally, how this CPS system can be implemented using standard technologies like OMG RTCORBA and DDS standard will be presented including a discussion on the necessary enhancements. We will also describe our test-bed to evaluate our solutions, which comprises a collection of open source wireless routers, netbooks, smartphones, and the BUG reconfigurable platform for embedded devices.
 

EMDS is a DDS-based data-centric multimedia content distribution approach. It eases the provision of new stream services because of the existing spatial-temporal decoupling between producers and consumers and because of it benefits from the great number of available DDS QoS policies. In EMDS, services such as stream transcoding, stream replication, stream composing, stream synchronization, device adaptation, and session recording (among others) can be developed and deployed on-line in a transparent manner with no client or protocol modifications. In the presentation we show how EMDS framework takes advantage of the data-centric approach and, more precisely how to use the built-in DDS discovery facilities and the DDS QoS policies for extensible, scalable and fault-tolerance on-line multimedia distribution services provision. A description of the design and implementation of DDS based multimedia services will be done. Also some use cases and scenarios will be shown. The presentation concludes with a discussion of the founded advantages and the lesson learned on using the proposed platform.
 

In systems that use primary-backup (passive) replication for fault tolerance, maintaining system availability after failures refers not just to ensuring the liveness of application functionality at a backup replica but also to ensuring that the state of the promoted backup matches that of the failed primary. A key requirement for implementing highly available services using passive replication is the ability to maintain consistency of application state across all the backup replicas of an application even in the presence of the dynamically changing operating conditions so that failure recovery is fast and predictable. However, this requirement adversely impacts response times perceived by client applications during the non-failure cases since the primary replica remains blocked until the state of all the backup replicas is made consistent with the state of the primary replica.

To overcome this limitation and to provide satisfactory response times for clients, a possibility is for the backup replica's state to be made consistent only intermittently or during failure recovery only, which significantly improves response times and saves resources, such as network bandwidth and CPU load, however, at the expense of a significantly weaker consistency model. This presentation describes how we are implementing such a tunable consistency mechanism within a real-time CORBA implementation, and how it aligns with the recent OMG standardization effort on lightweight real-time, fault-tolerance.
 

The US DoD GIG is a "globally interconnected, end-to-end set of information capabilities for collecting, processing, storing, disseminating, and managing information on demand to warfighters, policy makers, and support personnel". Physically this network expands worldwide and operates over shared and/or leased communication lines. At the edge of this System of Systems are subsystems dealing with real-time tasks such as situational awareness, command and control, and combat-management. This presentation will cover; Analysis of access control models with regards to their performance, scalability, deployability, and robustness characteristics as they apply to the real-time GIG environment; Analysis of available open-source and COTS technologies that can be used to describe access control policies, encode access rights, and enforce access control; Recent research on pragmatic access control models and security enforcement mechanisms suitable for highly dynamic real-time systems that can operate with high resilience in a decentralized environment; Proposed extensions to the DDS standard such that it can enforce access control and secure communications in these environments.
 

Wide-Area Situational Awareness (WASA) is recognized as a key enabling functionality for Smart Grids. It is one of the priority areas identified by Federal Energy Regulatory Commission (FERC) and by the National Institute of Standards and Technology (NIST) Smart Grid Interoperability Standards Effort. Time synchronized and precise grid measurements from Phasor Measurement Units (PMUs), called synchrophasor measurements, can provide a comprehensive view of the entire interconnection, when measurements from multiple utilities are combined, and improve wide-area situational awareness. Recognizing this potential, U.S. Department of Energy (DOE), North American Electric Reliability Corporation (NERC), and North American electric utilities, vendors, consultants, federal and private researchers and academics are collaborating on the North American SynchroPhasor Initiative (NASPI), whose vision is to improve power system reliability through wide-area measurement, monitoring and control. In this work, we present 1) the requirements for securing time-critical PMU data, 2) challenges in meeting those requirements and 3) some research directions towards addressing those challenges.
 

This presentation will introduce the Scala programming language and will make the case for it to become an important programming language for the future of computing and thus for OMG technologies such as DDS.
 

System designers and integrators in many industries are tasked with creating robust, widely available, and secure communications infrastructures. These infrastructures interconnect systems (including systems of systems) nationwide and globally with high fidelity and high performance.

Such a challenging real-time distributed system needs a well-designed, efficient architecture. The first step is to understand the challenges; this talk analyzes use cases common to defense, financial services, and power systems domains. The goal is to clarify interactions so we can next define system architecture and an efficient interface that meet the system's current needs and future expansion. The presentation will conclude with a proposed architecture based on OMG standards as well as a list of extensions to the standards that would benefit their use as the foundation for a large-scale data bus.
 

The concern that much of the developed nations' civil and military infrastructure may be vulnerable to cyber-attack has raised the security stakes for developers and integrators of critical-infrastructure systems. Security is on everyone's mind. While the over-arching application-level security goal is reasonably well understood, the requirements on the underlying middleware infrastructure are significantly less so. To what extent should middleware products and operating systems provide support for specific security approaches, offer plugins to integrate with external security services and components, or provide lower-level capabilities and assurances that are ultimately customized at the higher layer by application developers and integrators? What are the critical use-cases that illustrate the balance between open systems, information sharing, and access control? What are some of the immediate problems that must be solved? What are some of the fundamental engineering and deployment constraints? Our panel discusses the end-user and application-developer requirements imposed on network middleware systems with a specific focus on middleware that complies with the OMG Data-Distribution Service specification.
 

SESSION 7

Chair: Richard Warren, Principal Engineer, Real-Time Innovations
 

The growing set of OMG DDS specifications currently lacks support for data distribution over HTTP such that it is easy to access DDS pub/sub data from a lightweight client like a web browser, although this is the subject of the Web-Enabled DDS RFP currently in progress. This presentation explores the option of exposing a simple DDS API as a RESTful web-service. We explain the basic principles of REST and present a wish-list for the properties of a RESTful DDS service, as well as the DDS features it should support. This leads to an intuitive and simple RESTful DDS API, for which we present a prototype implementation.
 

In this talk we will present recent research on the integration of DDS with Enterprise Service Buses (ESBs), covering the architectural differences between the two integration models, the challenges and approaches to integrating DDS into an ESB, lessons learned from a prototype integration of DDS with Apache Camel and characterization of the more problematic scenarios, and benchmark results characterizing the impact on the scalability and performance of using an ESB-based approach versus a more native DDS approach to mediating DDS data.
 

System-Wide Information Management (SWIM) is an approach to the distribution of ATM (Air Traffic Management) information that seeks to address today's badly-integrated systems, which have evolved bottom-up from separate stakeholder sub-system and service-specific requirements. The FP6 SWIM-SUIT research project is prototyping a SWIM system by focusing on a subset of ATM stakeholders (and therefore requirements) which will have to be considered by the future SWIM. In this presentation we will show how the SWIM prototype has been designed and the results of the experiments that have been executed during the course of the project which involve distribution of large amount of flight data. Geographically distributed real word ATM applications and systems (e.g. Flight Data Processors, Flight Management Systems, Airport Management Systems etc) have been interconnected via the combination of request/reply and publish/subscribe patterns supported by the usage of novel technologies (for the ATM domain) such as web services and data distribution (DDS/JMS).
 

The rise in popularity of Cloud Computing and Cloud-based Service and the introduction of newly-defined web-based APIs to access those services has re-ignited the decade-old discussion between advocates of a REST/HTTP architecture for accessing services over the web and those promoting a Web-Services/WSDL/SOAP architecture. In the past this issue was of little relevance to real-time systems. However, the recent issuance of the Web-Enabled DDS RFP at OMG has changed all this: The resulting Web-API standard is expected to become a primary integration mechanism between enterprise and real-time (DDS-based) systems. This talk analyzes different REST API's used to access cloud computing resources, including the Amazon Cloud, the Open Cloud API, and the Kenai Project. 

It compares the practical differences and consequences of using a RESTful approach with a Web-Services Approach, and uses that as background to introduce recent research on mapping the data-centric DDS model to a RESTful API. Alternative mappings of DDS to REST are explored along with their scalability, security, usability, and performance implications.