Blockchain; A Game-Theoretic Cryptographic Ledger

Rasoul is an experienced scholar working interdisciplinary and multidisciplinary in computer science, and economics, and interested and well experienced in several topics such as multi- agent systems, game theory, economic mechanism design (social functions, double auctions, VCG mechanisms, matching theory, order books, etc.), decentralized economic.

In "cryptographic security", the security is a result of cryptographic primitives (hash functions, asymmetric cryptography, etc.) but in "game-theoretic security" the security is a consequence of disincentivizing malicious users. In this talk, we compare cryptographic protocol design and economic mechanism design. A cryptographic protocol is a finite sequence of messages between two or more agents. Cryptographic protocols are designed to provide us with security goals such as secrecy, authentication, key agreement, non-repudiation, anonymity, etc. An economic mechanism design is a multi-agent algorithm or a set of rules to incentivize and disincentivize rational (self- interest) individuals to follow certain policies. Economic mechanisms are designed to discover price in auctions, aggregate preferences in social choice functions, share resources based on Shapley value, match two sides of markets, etc., to provide fairness and efficiency regarding incentive compatibility. Blockchain is a decentralized ledger that works because of both cryptographic and game- theoretic security. We discuss consensus mechanisms from a game theoretic perspective regarding participants as rational players; how rational participants allocate their computational power? How they choose to continue a fork? How they select to join to a pool? How a pool manager shares the rewards among participants? etc. We introduce a set of axioms that a consensus mechanism should fulfill. We study the incentive compatibility of the Proof of Work consensus mechanism to moderate selfish mining, block withholding, pool hopping, etc. We also study the incentive compatibility of the Proof of Stake consensus mechanism for nothing-at-stake, grinding attack, etc.