DRD-EC: a dynamic risk-driven erasure coding method for IoT blockchains

Existing blockchain technologies applied in the Internet of Things (IoT) usually adopt the traditional full-replication data storage mechanism, which faces problems such as high storage overhead and poor scalability. In order to solve the above problems, we propose DRD-EC, a dynamic erasure coding storage method, which achieves an accurate assessment of node state and risk control by combining the node's reputation value and dynamic performance scoring and adopting an exponential decay model with adaptive weight adjustment. Meanwhile, according to the system risk and node state, the data redundancy and distribution strategy are dynamically adjusted, and a hybrid replication scheme is adopted, so as to reduce the storage overhead and ensure the read performance. Finally, a large number of experimental results illustrate that DRD-EC outperforms existing static erasure coding methods in terms of storage cost, fault tolerance, and read performance.