Clustering-Based Collaborative Storage for Blockchain in IoT Systems

Recently, blockchain is introduced to ensure the security of the device data in Internet of Things (IoT) systems. However, storing the entire blockchain ledger in resource-constrained IoT devices is impractical. A few existing papers attempt to mitigate the storage issues of the blockchain ledgers through the collaborative storage. Nonetheless, these studies solely consider collaborative storing of the entire blockchain ledger in a single consensus unit, treating all the nodes as an individual peer but ignoring which nodes should be assigned to form a consensus unit together. This may lead to significant latency in block invocations among the nodes. To this end, this article innovatively explores the grouping of all the devices in the IoT network into multipeers at a global level which significantly reduces the access latency. In this article, we propose a clustering-based collaborative storage scheme for the blockchain in storage-limited IoT systems. The proposed algorithm takes storage and communication latency into consideration and clusters various IoT nodes into multiple peers, ensuring that the entire blockchain stays updated within these clusters. Furthermore, we propose a series of effective block allocation and replacement strategies in both the static and dynamic scenarios. The experimental results verify that our algorithm effectively solves the problem of insufficient storage in blockchain systems.