The NetStack-RT Family

The TCP/IP protocols have been designed to operate over nearly any underlying local or wide area network technology. Although certain accommodations may need to be made, IP messages can be transported over a wide variety of physical layer and link layer technologies. However, with the proliferation of new technologies, there are ramifications of using the same design for every type of application and the efficiency of the TCP/IP stack varies for various applications. For many applications, the stack is proving to be a bottleneck due to various design and implementation constraints. Further, with the leaps being made in silicon speed and flexibility of hardware design, it is desirable to exploit the new features in embedded systems hardware while at the same time working transparently with networking applications written for a purely software network stack.

The NetStack/RT family of products is architected as an embedded solution from the ground up, and introduces new design breakthroughs to work around latency issues plaguing many implementations, using advanced buffer management & packet loss avoidance techniques and exploiting the fill power of multi-threaded operation.

The NetStack-RT Family comprises of:

  • Tech Specification

    NetF1 - High Octane TCP/IP

    NetF1 is a high performance, feature rich and hardware-acceleration capable embedded TCP/IP stack implementation with an included virtual routing framework. It includes a complete implementation of multi-instance capable TCP, UDP, IP, IPv6, ICMP, and IGMP designed specifically for use in low-resource embedded environments, where unused stack components can be scaled out or replaced with hardware equivalents. With full support for host and IPv4 / IPv6 router mode, NetF1 offers a superset of capabilities provided by many other host-only or router-only stacks including unnumbered interface support and interface identification for send/receive. This extensive feature-set does not come at the expense of performance either -- higher performance, binary compatibility with existing drivers and API-compatibility with existing socket based network applications are just a few reasons where NetF1 outpaces its peers. Further, it also includes the ability to optionally create highly multi-threaded multiple isolated and managed virtual "routers" in a single physical system thereby lending itself to use in embedded environments that need to separate out network backplane communication, or in ISP equipment that needs to support multiple routing protocols for multiple customers.

    Read more...
  • Tech Specification

    GrandPPPrix - PPP & Extensions

    GrandPPPrix is a robust, standards-based embedded implementation of the Point-to-Point Protocol (PPP), Multi-Link PPP (ML-PPP or MP), and Multi-Class Extensions to MP (MCMP). It includes the core components required to implement advanced, secure PPP in an embedded device to facilitate multi-protocol communication using single or multiple independent point-to-point links. With built-in support for prioritization, it also provides efficient transport of multimedia traffic and minimizes end-to-end delays. GrandPPPrix’s flexible compression, authentication, and encryption over diverse physical transports are ideal for use in constrained embedded environments.

    Read more...