A Good Horse Only Jumps as High as Needed: Multi-Field Fulcrum Codes for Heterogeneous Decoders

Abstract

For timely monitoring and effective decision making, the Internet of Things (IoT) requires reliable communications over the unreliable Internet, which is a very challenging task. FEC, such as RLNC and its variants, design coding schemes for a pre-defined set of devices that allow trading computation to achieve higher reliability with reasonable overhead and reduced latency. That, however, is ineffective for an IoT environment when devices are with diverse computing power. This paper introduces Multi-Field Fulcrum Network Coding (MF-FNC), allowing decoders of three different computation profiles to achieve high decoding probability. Simultaneously, our proposed MF-FNC reduces the computation complexity by up to 83%.