SecurShard: A Model for Hierarchical Fault Detection in Blockchain Sharding
Tirathraj Ramburn, Dhrubajyoti Goswami · 2023
Sharding increases concurrency in a blockchain system by splitting validators into groups (shards). Each shard processes a different transaction block such that throughput increases linearly in proportion to the number of shards. Contemporary sharding systems assume that shards are ‘perfect’. Therefore, shards need to be formed in such a way that they have negligible probability of failure. However, there are several limitations. In reality, shards can be faulty (violates the ‘perfect’ shard assumption) and an invalid block validated by a faulty shard cannot be detected until the block is appended to the blockchain. Thus, there is no fault detection mechanism during transaction processing. In this paper, we present a hierarchical fault detection model called SecurShard. The SecurShard model proposes mechanisms of combining shards into groups, one-to-all mapping of a transaction block to all shards in a group, and 100% consensus requirement to validate and append a block to blockchain; all these to ensure that a potentially invalid block is detected with high probability during transaction processing rather than after appending to blockchain. Furthermore, the mapping scheme relaxes the constraint that all shards need to be ‘perfect’. This also leads to the possible usage of smaller shards that are more performant than contemporary bigger shards while still maintaining the collective fault tolerance of the system. An elaborated theoretical analysis is presented which demonstrates the advantages of SecurShard in terms of fault tolerance and performance over traditional sharding systems.