Selected work

Representative Papers

Reserve Depletion and Security Runway in Proof-of-Stake Systems

Jun 02, 2026

This study investigates whether proof-of-stake (PoS) systems can securely transition to relying solely on transaction fees for security under finite token reserves. To this end, the authors develop a discrete-time stochastic model that captures validators’ strategic participation in response to price and demand fluctuations, characterize the equilibrium of their entry game, and introduce a state-dependent reserve threshold. This threshold partitions the system dynamics into three regimes: infeasible, reserve-dependent security, and pure fee-based security. By integrating a Markov decision framework with forward-looking participation conditions, the work overcomes the limitations of conventional reserve policies assessed merely by nominal exhaustion dates. The analysis yields explicit bounds on the probability of security failure and expected transition time, providing computable reserve requirements that enable quantitative stress testing of PoS systems’ capacity for secure fee-only operation.

0 citationsRead paper

On Sybil-proofness in Restaking Networks

Sep 22, 2025

This paper investigates the design of Sybil-resilient security mechanisms in re-staking networks. Addressing safety challenges arising from scaling validator responsibilities in existing protocols, it formally distinguishes— for the first time—two fundamentally distinct Sybil attack classes: *isolating* (where a single identity blocks others’ participation) and *colluding* (where multiple identities coordinate to launch attacks). It establishes an impossibility theorem: no universal slashing mechanism can simultaneously defend against both attack types. Leveraging random graph models—specifically Erdős–Rényi and bipartite block models—the paper analyzes the effectiveness boundaries of marginal and multiplicative slashing schemes, revealing that network homogeneity preserves Sybil resistance, whereas even minor heterogeneity drastically reduces attack costs. Key contributions include: (i) a rigorous classification framework for Sybil attacks; (ii) identification of an intrinsic limitation in incentive-compatible mechanism design; and (iii) quantitative characterization of how network topology decisively governs security guarantees.

0 citationsRead paper

Properties of UTxO Ledgers and Programs Implemented on Them

Jun 06, 2025Electronic Proceedings in Theoretical Computer Science

UTxO ledger models lack a formal trace semantics, hindering rigorous verification of smart contracts. Method: We model valid execution traces of UTxO smart contracts as infinite paths in the category of simple graphs and equip them with a hypermetric structure; we introduce an on-chain program representation based on non-expansive mappings, enabling rigorous embedding of stateful semantics into the stateless UTxO model. By integrating category theory, graph theory, topology, and trace semantics, we construct a unified analytical framework that formally relates execution traces to stateful program behavior. Contribution/Results: We prove key safety properties—including trace consistency and abstraction soundness—and establish a verifiable state abstraction mechanism. This work provides a solid mathematical foundation for high-assurance formal verification of UTxO-based smart contracts.

0 citationsRead paper

Enhancing Decentralization in Blockchain Decision-Making Through Quadratic Voting and Its Generalization

Apr 17, 2025

Blockchain governance suffers from excessive concentration of voting power and insufficient influence of small stakeholders. Method: This paper proposes Generalized Quadratic Voting (GQV), a mechanism employing three dynamic stake-allocation strategies to regulate decentralization. It integrates zk-SNARKs to ensure voter anonymity and verifiability. Contribution/Results: We formally prove that Type 2/3 QV and GQV significantly improve both the Gini coefficient and Nakamoto coefficient—quantitative measures of power distribution and consensus centralization. A tunable threshold parameter enables on-demand customization of decentralization levels. We identify and characterize a critical “influence leap” phenomenon among marginal shareholders under QV. Empirical evaluation demonstrates GQV’s effectiveness in DAO governance, community collaboration, and public-good decision-making: it substantially reduces voting power concentration while preserving fairness and amplifying the voice of small- and medium-sized participants.

0 citationsRead paper

Recent publications

Latest Papers

Reserve Depletion and Security Runway in Proof-of-Stake Systems

Jun 02, 2026

This study investigates whether proof-of-stake (PoS) systems can securely transition to relying solely on transaction fees for security under finite token reserves. To this end, the authors develop a discrete-time stochastic model that captures validators’ strategic participation in response to price and demand fluctuations, characterize the equilibrium of their entry game, and introduce a state-dependent reserve threshold. This threshold partitions the system dynamics into three regimes: infeasible, reserve-dependent security, and pure fee-based security. By integrating a Markov decision framework with forward-looking participation conditions, the work overcomes the limitations of conventional reserve policies assessed merely by nominal exhaustion dates. The analysis yields explicit bounds on the probability of security failure and expected transition time, providing computable reserve requirements that enable quantitative stress testing of PoS systems’ capacity for secure fee-only operation.

0 citationsRead paper

On Sybil-proofness in Restaking Networks

Sep 22, 2025

This paper investigates the design of Sybil-resilient security mechanisms in re-staking networks. Addressing safety challenges arising from scaling validator responsibilities in existing protocols, it formally distinguishes— for the first time—two fundamentally distinct Sybil attack classes: *isolating* (where a single identity blocks others’ participation) and *colluding* (where multiple identities coordinate to launch attacks). It establishes an impossibility theorem: no universal slashing mechanism can simultaneously defend against both attack types. Leveraging random graph models—specifically Erdős–Rényi and bipartite block models—the paper analyzes the effectiveness boundaries of marginal and multiplicative slashing schemes, revealing that network homogeneity preserves Sybil resistance, whereas even minor heterogeneity drastically reduces attack costs. Key contributions include: (i) a rigorous classification framework for Sybil attacks; (ii) identification of an intrinsic limitation in incentive-compatible mechanism design; and (iii) quantitative characterization of how network topology decisively governs security guarantees.

0 citationsRead paper

Properties of UTxO Ledgers and Programs Implemented on Them

Jun 06, 2025Electronic Proceedings in Theoretical Computer Science

UTxO ledger models lack a formal trace semantics, hindering rigorous verification of smart contracts. Method: We model valid execution traces of UTxO smart contracts as infinite paths in the category of simple graphs and equip them with a hypermetric structure; we introduce an on-chain program representation based on non-expansive mappings, enabling rigorous embedding of stateful semantics into the stateless UTxO model. By integrating category theory, graph theory, topology, and trace semantics, we construct a unified analytical framework that formally relates execution traces to stateful program behavior. Contribution/Results: We prove key safety properties—including trace consistency and abstraction soundness—and establish a verifiable state abstraction mechanism. This work provides a solid mathematical foundation for high-assurance formal verification of UTxO-based smart contracts.

0 citationsRead paper

Enhancing Decentralization in Blockchain Decision-Making Through Quadratic Voting and Its Generalization

Apr 17, 2025

Blockchain governance suffers from excessive concentration of voting power and insufficient influence of small stakeholders. Method: This paper proposes Generalized Quadratic Voting (GQV), a mechanism employing three dynamic stake-allocation strategies to regulate decentralization. It integrates zk-SNARKs to ensure voter anonymity and verifiability. Contribution/Results: We formally prove that Type 2/3 QV and GQV significantly improve both the Gini coefficient and Nakamoto coefficient—quantitative measures of power distribution and consensus centralization. A tunable threshold parameter enables on-demand customization of decentralization levels. We identify and characterize a critical “influence leap” phenomenon among marginal shareholders under QV. Empirical evaluation demonstrates GQV’s effectiveness in DAO governance, community collaboration, and public-good decision-making: it substantially reduces voting power concentration while preserving fairness and amplifying the voice of small- and medium-sized participants.

0 citationsRead paper