Capillary routing for reliable real-time streaming

Application of forward error correction (FEC) in offline streaming with large buffering time dramatically improves the quality and performance of communications in challenging network conditions. However real-time streaming puts hard restrictions on the buffer size leaving FEC helpless for combating long link failures on a single path route. Multi-path routing is another method, orthogonal to buffering, which can make FEC effective also for real-time streaming. In this paper we introduce a capillary routing algorithm offering layer by layer a wide range of multi-path routing topologies starting from a simple multi-path solution toward more reliable and secure schemes obtained by recursively spreading individual subflows. The friendliness of a particular multi-path routing scheme is rated by a measure called Failure Recovery Redundancy Overall Requirement (FRROR), which is proportional to the total amount of FEC codes required for combating the individual failures of all links in the multipath route. A dozen of capillary routing layers, built on several hundreds of network samples obtained from a random walk wireless Mobile Ad-Hoc Network (MANET), are rated with FRROR. We show that the overall requirement in FEC decreases substantially as the spreading of the routing grows.