Bayesian Mechanism Design for Blockchain Transaction Fee Allocation

TL;DR

The paper introduces a BNIC-based transaction fee mechanism design for blockchains by developing an auxiliary mechanism method that links BNIC with DSIC frameworks. Using a multinomial logit allocation model, it constructs a BNIC and 1-SCP TFMs that yield a positive miner revenue, achieving a constant-factor approximation to the optimal revenue under i.i.d. bounded valuations, and extends the approach from block size 1 to general $k$ with a weighted-sampling allocation. It proves an almost miner-incentive-compatibility (MIC) result, and establishes almost-MIR properties with revenue concentration, while presenting an impossibility result showing that deterministic BNIC+1-SCP mechanisms cannot guarantee positive miner revenue. The work also discusses cryptographic on-chain implementations and compares to EIP-1559 and MPC-based alternatives, highlighting practical implications for incentive alignment, collusion resistance, and revenue stability in decentralized systems.

Abstract

In blockchain systems, the design of transaction fee mechanisms is essential for stability and satisfaction for both miners and users. A recent work has proven the impossibility of collusion-proof mechanisms that achieve both non-zero miner revenue and Dominating-Strategy-Incentive-Compatible (DSIC) for users. However, a positive miner revenue is important in practice to motivate miners. To address this challenge, we consider a Bayesian game setting and relax the DSIC requirement for users to Bayesian-Nash-Incentive-Compatibility (BNIC). In particular, we propose an auxiliary mechanism method that makes connections between BNIC and DSIC mechanisms. With the auxiliary mechanism method, we design a transaction fee mechanism (TFM) based on the multinomial logit (MNL) choice model, and prove that the TFM has both BNIC and collusion-proof properties with an asymptotic constant-factor approximation of optimal miner revenue for i.i.d. bounded valuations. Our result breaks the zero-revenue barrier while preserving truthfulness and collusion-proof properties.

Figures (5)

• Figure 1: The role of different parties in generating one block in a blockchain.
• Figure 2: Illustration of the blockchain structure. Above the dashed lines are the blocks and their contents (transactions and rewards). The blocks are arranged from left to right in the time order, and each block is linked to the previous one on its left. Below the dashed lines are the states of the system -- the amount of money owned by each party at the time.
• Figure 3: The plot of $g(\lambda_0)$ in Theorem \ref{['thm:h:value:k']}.
• Figure 4: The diminishing probability of MIR being violated for the uniform distribution $b_i\sim \mathrm{Unif}[0,1]$.
• Figure EC.1: The plot of $f(\cdot)$.

Theorems & Definitions (35)

• Theorem 1: shi
• Definition 1: Transaction Fee Mechanism
• Definition 2: Symmetry
• Definition 3: User Dominant-Strategy-Incentive-Compatibility (U-DSIC)
• Definition 4: User Bayesian-Nash-Incentive-Compatibility (U-BNIC)
• Definition 5: User Sybil-Proofness (U-SP)
• Definition 6: $c$-Side-Contract-Proofness ($c$-SCP)
• Lemma 1: Myerson's Lemma myerson1981optimal
• Definition 7: Auxiliary-Variation Decomposition
• Definition 8