SDN-based Optimizing Solutions for Multipath Data Transmission Supporting Consortium Blockchains

Abstract

The user can organize a social network together with other peoples, in the manner of the consortium blockchain, so that expected contents can be accessed via collecting pieces of data (sub-flows) from members. This promising mechanism heavily relies on the multipath data transmission to transfer sub-flows to the user over the social network. However, the performance differentiation and link overlapping of transmission paths between the user and members could result in severe traffic congestion and packet scrambling. To solve these problems, we design a novel SDN (Software-Defined Networking)-based framework because the global traffic engineering can perform more flexible flow transferring. Our SDN controller has three modules of topology discovery, sub-flow association and path selection. We quickly capture the social network from the most appropriate consortium blockchain where a set of paths with minimal link overlapping and performance differentiation can be found to transfer correlated sub-flows the user wants. The simulation results show that our solution achieves a higher network throughput compared with traditional methods while guaranteeing a low data retransmission rate.