A Proposed Permissioned Blockchain Consensus Algorithm: Consensus Algorithm Genetically Enhanced (CAGE)

Blockchain technology is disruptive and relies on the development of consensus algorithms, CAs. Currently, there are many CAs proposed and this paper explores a new avenue of nature-inspired CAs. In particular, emphasis is on selection techniques used in canonical genetic algorithms, GAs, combined with a stamina attribute. The premise is that selection techniques used in GAs could be adapted and used to select nodes responsible for adding blocks in permissioned blockchain networks. Various selection techniques were researched and the Fitness Proportionate Selection was chosen as the most suitable, which was tested on a blockchain network and executed many times and compared to another CA for permission blockchain networks, Proof-of-Elapsed-Time, PoET. The results of the experiments show an improvement in transaction throughput and block creation when compared with PoET. Our initial results, however, had a biased node selection and could be vulnerable to take-over attacks. Therefore, a stamina element was introduced and the experiments were repeated with the results showing a less biased node selection. The overall contribution is the exploration of nature as a solution to CAs. The initial results are comparable and in places better than existing CAs, such as PoET.