On a Class of Multi-Source Distributed Storage With Exact Repair

Abstract

In future communication networks, say 5G networks and Internet of Things, users may need to obtain messages from multiple available sources in a distributed manner, which requires distributed storage in the network. One essential issue in distributed storage is how to repair a failed storage node. In conventional distributed storage model with exact repair, source files are stored in many nodes in a distributed manner and the contents of a failed node are exactly recovered by the surviving nodes. Further, all files can be reconstructed when a decoder has access to a certain number of storage nodes. So, the underlying reconstruction network is equivalent to a single-source multicast problem. This paper considers a variant of the exact repair problem, where the underlying reconstruction network is the independent distributed source coding system, a type of multi-source multicast problem. As the first non-trivial case with two sources and three encoders, the storage-repair tradeoff regions are proved for all the 33 non-isomorphic instances, and it is shown that binary codes are optimal.