A survey on standards for real-time distribution middleware

Abstract

This survey covers distribution standards oriented to the development of distributed real-time systems. Currently, there are many distribution standards that provide a wide and different set of real-time facilities to control the temporal aspects of applications. Besides giving a general overview of these standards, we describe the real-time mechanisms proposed by each standard to manage both processor and network resources, discuss whether the available facilities are sufficient to guarantee determinism throughout the whole application, and identify a set of features and deployment options that would be desirable in any real-time distribution middleware regardless of its distribution model and standard. The survey identifies open issues and key challenges for future research.

An IoT Middleware Framework for Industrial Applications

“…A very interesting discussion on the utilization of standards for real-time distribution middleware is presented in [31]. The authors, out of several distribution models, chose those which are based on the standard, are mature, stable and with impact on the industry; namely: CORBA/RT-CORBA, Distributed System Ada Annex (DSA), Data Distribution Service for Realtime System (DDS) and Distributed Real-Time Specification for Java (DRTSJ).…”

“…The authors, out of several distribution models, chose those which are based on the standard, are mature, stable and with impact on the industry; namely: CORBA/RT-CORBA, Distributed System Ada Annex (DSA), Data Distribution Service for Realtime System (DDS) and Distributed Real-Time Specification for Java (DRTSJ). Even though the authors of [31] do not provide a verdict or have not carried out a ranking, however, a classification can be made.…”

“…CORBA/RT-CORBA has the following advantages: it is based on a very mature technology, one involved in a wide range of applications [31], such as Software Defined Radios [32] and Industrial Robotics [33]; RT-CORBA entities validate 33 | P a g e www.ijacsa.thesai.org the development of critical real-time applications; from the point of view of scheduling, the RT-CORBA provides static scheduling based on Fixed Priority Scheduling (FPS), the use of threads as schedulable entities, control of the competition degree on the servers using thread pools, deterministic access to shared resources, the use of different scheduling policies, and the use of distributable threads as a schedulable entity; as far as network resource management is concerned, it provides mechanism for the fine-tuning of network properties, it uses private connections and definitions of priority-banded connections; it is the only standard which provides mechanisms for the specification of scheduling parameters which may be used during execution; facilitate interoperability between implementations (GIOP -General Inter-ORB Protocol); TAO implementation is the most popular and updated open-source implementation for RT-CORBA. As disadvantages of RT-CORBA, we can mention: unlike the CORBA specification updated in [34], RT-CORBA is not currently in the attention of the Object Management Group (OMG), the last update being performed in 2005 [35][36]; it does not take into consideration the network scheduling; it uses TCP/IP stack which means that even the use of Ethernet switches is unsuitable for implementation of hard real-time systems; TAO implementation does not provide synchronous protocols (it is based on the operating system); it does not implement the priority transforms model, the use of buffers to store remote requests in thread pools nor the borrowing of threads among thread pool lanes.…”

“…The DDS has the following advantages: it is considered a mature technology involved in several real-time applications [31] in the fields such as Defence [17], Automation [37], and Space [38]; supports anonymous and asynchronous dissemination of information; has specific requirements for distributed applications such as control systems, sensor networks, and industrial automation systems; it is a data-centric middleware [18] and, therefore, it is aware of the contents of the interchanged data which can be directly managed; it provides multiplatform and multi-language support; the types of shared data can be defined by using IDL language [34]; interoperability between different implementations is provided by DDS Interoperability Wire Protocol (DDSI) [39]; it is a recently updated specification, OMG provides specification for the Extensible and Dynamic Topic Types [40], which provides support in order to define and modify dynamic (on runtime) data for the extension and evolution of systems based on DDS; the DDS model defines a strongly typed Global Data Space where publishers (Data Writer (DW)) can write (provide) data and subscribers (Data Reader (DR)) can read (consume) data allowing the middleware to focus on obtaining data independent of their origin; the standard was explicitly designed for distributed real-time systems; specifications define a set of QoS parameters in order to configure nonfunctional properties for each entity and allow the change of some of them during an operation; a subset of QoS parameters allows the control of temporal behaviour and improves the application predictability; it defines different mechanisms meant to validate the communication between entities (polling, synchronous mode and asynchronous mode for the DR entity) and provides the opportunity to notify the application by Polling, Listeners, Conditions, and Wait-sets; there are both commercial (CoreDX or RTI-DDS) and open source (OpenSplice or OpenDDS) implementations. Among the DDS disadvantages, we can mention: there are no evaluations in detail done on the DDS real-time performance (an attempt can be found in [41]); it does not explicitly addressed the scheduling of threads at processor level; it is oriented on IP networks and not on the real-time networks (still lists a set of requirements for network support); considers only network policies based on fixed priority scheduling and excludes any other type of predictable network used in industry; some internal middleware operations generate meta-traffic thus introducing an override that must be taken into account in the analysis of behaviour in time; DDSI has an indefinite number of sub-messages; there is still no profile for safety-critical applications.…”

MCIP Client Application for SCADA in Iiot Environment

“…The mobile revolution, with more than 5 billion smart phones connected to various mobile networks, provides the population with the possibility to access the Internet. www.ijacsa.thesai.org DDS RTS (Data Distribution Service for Real-Time Systems) is a middleware standard based on the publish/subscribe paradigm for distributing data between heterogeneous applications developed by OMG consortium [12]. An important feature of this protocol is that it has facilities for implementing QoS (Quality of Service) parameters in order to achieve real-time performance.…”