The recently proposed Transaction Fee Mechanism (TFM) literature studies the strategic interaction between the miner of a block and the transaction creators (or users) in a blockchain. In a TFM, the miner includes transactions that maximize its utility while users submit fees for a slot in the block. The existing TFM literature focuses on satisfying standard incentive properties – which may limit widespread adoption. We argue that a TFM is “fair” to the transaction creators if it satisfies specific notions, namely Zero-fee Transaction Inclusion and Monotonicity. First, we prove that one generally cannot ensure both these properties and prevent a miner’s strategic manipulation. We also show that existing TFMs either do not satisfy these notions or do so at a high cost to the miners’ utility. As such, we introduce a novel TFM using on-chain randomness – rFTM. We prove that rFTM guarantees incentive compatibility for miners and users while satisfying our novel fairness constraints.

Rational Consensus (RC) is a more realistic modelling of the traditional Byzantine Consensus problem, motivated by the recent works in Rational Cryptography. RC is the problem of achieving consensus in the presence of Rational, Byzantine and Honest players (players- participants in the consensus protocol) in a distributed system. This work focuses on consensus in multiple rounds with additional agreement on ordering among rounds which is a more general problem called Atomic BroadCast (ABC). Blockchains is an example of application of ABC. This work abstracts rational players in three types on their incentive structure. We show the impossibility of achieving consensus for two out of the three types of rational players under some conditions. For the third type of rational players, existing work models a single round of agreement and, therefore, doesn't capture the existence of another insecure equilibrium strategy for rational players. We finally fill the gap in the literature of a Rational ABC by proposing a novel protocol for rational consensus, namely pRFT. We prove (i) the correctness of the protocol and (ii) the communication complexity of pRFT, which is a form of accountable protocol, equals the best-known accountable agreement protocols.

Proof-of-Stake(PoS) based blockchain protocols have gained popularity due to their higher throughput and low carbon footprint when compared with Proof-of-Work blockchain protocols. The two major parts of blockchain protocols are the selection of the next block proposer and the selection of the longest chain. In PoS the block publishers are selected based on their relative stake. However, PoS-based blockchain protocols may face vulnerability against Fully Adaptive Corruptions. This paper proposes a novel PoS-based blockchain protocol, QuickSync, to achieve security against Fully Adaptive Corruptions while improving performance. Towards this, we propose a metric for each block: block power. We compute the chain power of a chain as the sum of block powers of all the blocks comprising the chain. The chain selection rule selects the chain with the highest chain power as the valid chain. Since the block proposer is not selected upfront, this scheme is resilient to fully adaptive corruptions, which we also show formally. We also ensure that our block power mechanism is resistant to Sybil attacks. We prove the security of QuickSync by showing that it satisfies the common prefix, chain growth, and chain quality properties. Our analysis demonstrates that QuickSync performs better than Bitcoin by order of magnitude on both transactions per second and time to finality.