Crypto & Blockchain · Thu, Nov 12, 2026

Bitcoin, Ethereum, stablecoins, DeFi, smart contracts, scaling and consensus.

RSS feed
arXiv cs.CR · cs.GT · q-fin.TR

Verifiable Settlement and Pricing for Local Energy Markets: Architecture and Deployment on Hyperledger Besu

In briefVerifiable Settlement and Pricing for Local Energy Markets on Hyperledger Besu

SummaryThis paper implements privacy-preserving local electricity trading on a Hyperledger Besu network with QBFT consensus, using an automated market maker to set prices. Individual electricity positions stay off-chain, while only aggregate supply and demand are published, and smart contracts compute prices every fifteen minutes. Poseidon2 commitments and UltraHonk zero-knowledge proofs generated from Noir circuits verify aggregation, payments, and daily balance updates. In a ten-household simulation, local trading covered 34.4% of demand and reduced grid-only costs by 30.5% under assumed tariffs.

Michele Fabi (CREST), Baptiste Leloup (ENSAE)

★★★★★ Notable
arXiv cs.CR · cs.NI

What Does It Cost to Force-Close a Lightning Channel? Evidence from 974,000 Channel Closes, 2018-2026

In briefCost of force-closing a Lightning channel: analysis of 974,000 channel closes, 2018–2026

SummaryThis paper traces 973,581 Lightning Network channel closes from January 2018 to October 2026 on-chain and cross-checks 698,476 public closes with a separate chain index. It finds that fee spikes did not produce a wave of force closes. Force closes cost two to six times as much in fees as cooperative closes.

WhyThe authors measure force-close costs across fee regimes using 973,581 channel closes, informing zero-fee commitment and anchor-output design decisions.

Saurabh Kumar

arXiv cs.LG · cs.AI · cs.CR

Tokenized Reward Systems for Federated Learning: The Privacy vs. Fairness Tradeoff

In briefBlockchain reward systems for federated learning and the privacy-fairness tradeoff

SummaryThis paper designs and tests two Solidity smart contracts that distribute rewards in federated learning, one based on a public quality score and one that hides scores using a zero-knowledge proof. Experiments on MNIST with Sybil attacks, zero-gradient updates, and random noise show that the privacy-preserving version is less robust. Adversaries capture roughly double the average reward, and honest contributors can be fully excluded in later training stages.

Andreas Sendros, Nikolaos Pavlidis, Vasilios Perifanis et al.

arXiv cs.DC · cs.CR

SoK: Design Choices in Tendermint BFT Consensus

In briefA systematic survey of design choices in Tendermint BFT consensus

SummaryThis paper surveys the Tendermint Byzantine fault-tolerant consensus protocol and the research that proposes changes to it. It describes the baseline protocol by its design choices, such as vote dissemination and proposer rotation, and classifies each work by the choice it modifies and the properties it strengthens or weakens. The surveyed solutions are grouped into three concerns: performance, validator accountability, and protection of exploitable information. The authors also identify recurring strategies and pose open questions for future work.

Sanaz Taheri-Boshrooyeh, Yahya Hassanzadeh-Nazarabadi