RIPlink technology

A:  RIPlink stands for "Remote IP link" and extends IP networks' reach to limited resource devices connected on non-IP networks. Very simple devices (sensors for example) can run a TCP application and connect to IP over its low bandwidth network with minimum impacts.

A:  The device sends/receives over the non-IP network its IP application data to/from a specific gateway, using the simple Remote Socket Protocol (RSP). The gateway, running the IP stack and connected to IP network, transfers all IP data between the device and the IP network. The device uses simple standard TCP/UDP application programming interface, while the IP complexity is handled by the gateway. See also RIPlink Technical Overview

A:  From technological perspective, there is nothing new. It is using well known Remote Procedure Call (RPC) principle for well-known TCP/UDP sockets. And this has already been proposed by some university work in the past.
What is new is having a standardization approach to make this technology widely adopted and interoperable, in particular applied to machine-to-machine applications.

A:  The main benefits are those of IP, applied to all type of networks and devices. IP benefits are to make application design network independent, portable, standard and interoperable. With RIPlink, IP benefits apply to lightweight devices (sensors, resource limited machines, etc.) and limited wired and wireless networks (low-power RF such as ZigBee?, low bandwidth power line communication, CAN, etc. ). RIPlink is simple, i.e. easy to develop and deploy.

A:  By its principle, IP addressing operates at application level, not at network level. The device IP application is addressed through a port of the gateway IP address. This is similar to a NAT approach, where the device itself is not visible on the IP network.

A:  The objective are similar, but the approach is different: while 6lowPAN and ROLL define an adapted IP network layer (in particular routing) for low power lossy networks, RIPlink operates simply at application level: underlying non-IP network and routing is unchanged, and IP routing is not impacted. 6lowPAN and ROLL benefits is to have IPv6 address assigned to each device, but with the counterpart of more resources on the device (memory, CPU and bandwidth - thus more power consumption), no direct portability on existing network technologies, and still a proxy between the standard IPv6 and the 6lowPAN network.

A:  It is perfectly suited to integration of new applications on resource limited devices, especially when combined with heterogeneity of application, device and networks. Machine-to-machine applications can benefit a lot of this technology, by making sure same application can be ported on different type of specialised wired and wireless network, on different devices possibly running on battery. Typically, smart metering, alarm and security, home, building and industrial automation will benefit from RIPlink for introducing, faster and safer, new applications and services. See some typical applications.

A:  Yes. Whatever the type (wired or wireless), the performances (bandwidth, latency), the maximum packet size, RIPlink can be easily adapted. When existing, the network built-in mechanism for acknowledgement or in-order packet delivery will be re-used, otherwise they will be added.

A:  Yes. The constraints on the device are very limited: typically few kilobytes of code (<10kB), very low data memory, no significant CPU requirement, no real-time constraints, no OS required. It is perfectly suited to devices running on battery, with very limited CPU.

A:  The gateway has to support one or more non-IP network interface and to implement the RSP protocol. It needs TCP/UDP/IP stacks and IP network interface(s), except in the case RIPlink is used locally as a stand-alone - not connected to IP - system.

A:  The Remote Socket Protocol (RSP) is the client-server protocol used between the devices and the gateway to transport IP application connection request and data. It is standardised to ensure interoperability.

A:  RIPlink is built as a standard: specifications are made freely available to ensure that RIPlink can be implemented by any one, in an interoperable manner. There is not yet a formal RIPlink certification process in place. By adopting RIPlink, you will participate to turning it into an industry standard.

A:  An evaluation package is available for download to evaluate the technology. Such package is modular and suited to various needs: understanding the technology, testing an end-to-end application, testing a new non-IP network interface, testing a gateway, etc.