AssociateChain: Scaling blockchain in cloud–edge-enabled Metaverse via associative sharding

Abstract

Blockchain has been extensively employed to strengthen the security and privacy of cloud–edge-enabled Metaverse due to its immutability, decentralization, and other key characteristics. However, limited scalability remains a significant barrier to its broader adoption. Sharding offers a promising solution for scaling blockchain, but existing schemes perform state and node sharding separately, ignoring the fact that, in cloud–edge computing, end devices are typically serviced by the nearest edge nodes to achieve a high quality of service (QoS). We define this characteristic as device–edge proximity. As a result, device states may be assigned to distant edge node shards, requiring transactions to be relayed over the wide-area network before reaching their assigned shards, thus introducing additional relay overhead and increasing the overall transaction latency. To address this challenge, we propose AssociateChain, which employs a two-stage associative sharding approach to minimize number of relay transactions. In the first stage, device states are grouped into shards based on historical transaction patterns. In the second stage, the edge node sharding process is modeled as a stable matching problem, optimizing the assignment of edge nodes based on the first-stage results and device–edge proximity. Furthermore, we introduce a double-check mechanism to strengthen resilience against potential takeover attacks. Although AssociateChain slightly increases sharding complexity from $O\left(n\right)$ to $O\left(k(n+1)\right)$, where n is the number of nodes and k is the number of shards, experiments demonstrate that it reduces the relay transaction ratio by 90%, lowers relay delay by two orders of magnitude, and decreases total cost by 70%, significantly outperforming existing sharding schemes.