FAQ’s
Q. I’ve never heard of FEC/ECC, what is it?
A. Forward Erasure Correction (FEC) or using Erasure Correcting Codes (ECC) in general provide a mechanism to protect data from the effects of packet loss by creating repair packets in advance which can be used to recreate data which is lost. Extensively used in storage systems, now increasingly also used in data transport systems.
Q. What’s special about Steinwurf’s FEC /ECC products?
A. Steinwurf’s FEC / ECC products utilise the latest innovations in coding Technology called Random Linear Network Coding (RLNC), a powerful new technology that can be used to improve the performance of many of today’s communication systems. This allows our FEC solutions to operate in ways other codes cannot, in new and entirely flexible ways which can adapt to changing traffic and network conditions. Essentially you can get the benefit of all that historically used codes can do, with a number of new features that no other codes can match.
Q. What is RLNC?
A. RLNC was developed across MIT, Caltech and over 10 other institutions, providing a way to practically implement Network Coding. Designed to address system level inefficiencies and to be backward compatible with prior generations of coding technologies, RLNC finally allows communications to move away from an end to end operation, and be optimised also at a much more granular network level, among and between nodes in the network.
Q. So are Steinwurf’s FEC solutions block, rateless, fountain codes or something else?
A. Steinwurf’s modern erasure codes are network codes, and have the flexibility and ability to operate as any of these types of codes. Uniquely, they also make it possible to operate as sliding window codes.
Q. Where can Steinwurf’s FEC be used?
A. Steinwurf’s FEC can help enable low latency communications; for multicasting or broadcasting data to thousands of devices at once; to speed up communications over the internet, improve video quality for streaming movies and live events; make satellite comms more efficient; and optimize data storage size while increasing reliability for the datacenters used for cloud computing. Anywhere there are inefficiencies and packet losses, especially in wireless systems, Steinwurf can help.
Q. How easy is Steinwurf’s FEC to implement?
A. Depending on your requirements, Steinwurf’s FEC can be implemented in any or multiple layers of the network stack. From a system point of view it should be implemented end-to-end. Often the simplest method is at the application layer, with a software update. All that is required is for a source device to be loaded with our software to encode data, and for a destination device to be loaded with our software to decode the data. It’s possible to go further and also create systems with intermediary nodes containing our software to ‘recode’ data between the source and destination, for maximum efficiency along any path.
Q. What protocols can be used between the encoder and decoder?
A. In general we suggest one of two types of implementation. Where the existing software uses an unreliable datagram oriented transport (such as UDP). This could be a video streaming application. Can be directly integrated into the existing software’s networking layer. Where the existing software uses a connection oriented transport (such as TCP), the integration needs to happen below TCP, in a tunnel perhaps, similar to how a VPN works.
Q. Do these FEC codes require a lot of computing power?
A. All Steinwurf’s algorithms are optimised for hardware acceleration, and on modern cpus the resource cost is barely perceptible. See our blog for further details: https://www.steinwurf.com/blog/hardware-acceleration Of course if you talk to us about your use case and desired implementation, with our knowledge and reputation as the go-to FEC Experts, we can help you deploy our technology in the most efficient manner for your particular use case.
Did we answer your question here?
If not, get in touch to learn more from Steinwurf - The FEC Experts
contact@steinwurf.com