Real-Time Innovations (RTI), The Real-Time Middleware Experts, announced support for lossy networks in RTI Data Distribution Service 4.3. Lossy networks — those that have uncertain or sporadic connectivity — are common in a variety of applications, such as defense, unmanned vehicles, oil production and other resource-management applications.
Most distributed systems today rely on messaging middleware that uses the Transmission Control Protocol (TCP) for inter-application and inter-node communication. TCP is popular because it provides reliable delivery of messages and data. However, TCP has characteristics that make it undesirable for use in real-time applications where the underlying network itself is not reliable.
RTI provides a built-in transport that is Internet Protocol (IP)-based and employs the User Datagram Protocol (UDP). RTI also provides a completely tunable reliability model to optimize transport utilization over transient, high-delay, bandwidth-limited and lossy networks.
“RTI’s support for reliable messaging and data distribution over unreliable networks gives developers an off-the-shelf, standards-compliant alternative to developing proprietary communicationn software,” said David Barnett, vice president of Marketing at RTI. “RTI can be used in systems that are deployed in challenging, non-enterprise environments, such as the battlefield, air and space. These applications and networks are not supported by traditional enterprise messaging and integration middleware.”
One of the many ways the RTI reliability model provides higher utilization of low-bandwidth networks is that it allows the frequency of heartbeats and acknowledgements to be fine-tuned. The result is efficient and reliable transport for unreliable networks that can be tuned as network conditions and bandwidth requirements fluctuate. In environments that typically experience up to three percent packet loss, RTI can achieve greater than 90 percent network utilization, which is far superior to TCP performance.
The RTI reliability model is completely configurable, enabling developers to achieve the appropriate balance of determinism and reliability even in the presence of bandwidth-limited, high-latency and lossy transports. The use of UDP in conjunction with the RTI reliability model supports features not available with TCP, such as multicast, for extremely efficient data distribution.
About RTI Data Distribution Service
RTI Data Distribution Service is a high-performance messaging and data-caching solution for the development and integration of applications that require low latency, high throughput, high scalability, deterministic responses and minimal consumption of network, processor and memory resources. RTI Data Distribution Service meets the unique and demanding requirements of mission-critical real-time systems, including the ability to run in dynamic and autonomous environments and over unreliable or low-bandwidth networks such as wireless and satellite links.
RTI Data Distribution Service complies with the Object Management Group (OMG) Data Distribution Service for Real-Time Systems (DDS) standard. RTI is the world’s leading supplier of DDS-related software and services, with more than four times the market share of all other DDS suppliers combined.
Real-Time Innovations (RTI) provides high-performance infrastructure solutions for the development, deployment and integration of real time, data-driven applications. RTI’s messaging, caching, Complex Event Processing (CEP) and visualization capabilities deliver dramatic improvements in latency, throughput and scalability while slashing cost of ownership. The company’s software and design expertise have been leveraged in a broad range of industries including defense, intelligence, simulation, industrial control, transportation, finance, medical and communications. Founded in 1991, RTI is privately held and headquartered in Sunnyvale, CA.
RTI, Real-Time Innovations, The Real-Time Middleware Company and RTI Data Distribution Service are registered trademarks or trademarks of Real-Time Innovations, Inc.