A major technical challenge of the cryptocurrencies is to find a way to safely increase the throughput of the system in terms of number of transactions. An approach to tackle this limitation is to review the role of the blockchain, or even to take that data structure out of the picture completely. In this post, we will comment a paper by Boyen, Carr and Haines named Blockchain…

Log-Structured-Merge trees (LSMT) are a good fit for modern SSD storage and offer good performance and reliability. LSMT are also a good fit for blockchain storage requirements (snapshots, consistency, proof of existence). This blog post describes a database designed specifically for blockchain storage, inspired by existing LSMT implementations (RocksDB, COLA tree).

The current state-of-the-art LSMT implementation is probably RocksDB, with in-memory write buffers, parallel compaction and snapshots. Another similar algorithm are COLA tree. That is a btree-like structure where each node has separate write buffer. Finally there is SSTable from Cassandra…

This post tries to give a short overview of provable security in cryptocurrencies.

For example, the paper "New Directions in Cryptography" by Whitfield Diffie and Martin Hellman, which is considered by most the beginning of modern cryptography (at least the public and civilian one), does not provide such rigorous analysis.

The publications from the cryptographic research…

Bernardo, the presenter, divided the talk in two parts: the first reviews main topics in Cryptography which would help the viewer to understand the presentation and the protocol itself. Whereas the second is about the protocol itself.

First Part - Cryptography background

• Commitments
• Coin Tossing/Guaranteed Output Delivery
• Verifiable Secret Sharing

Second Part - Proof-of-Stake…

In this article I'm going to provide a brief review of protection methods against replay attacks, arising from signature malleability of elliptic curve cryptography.

Thus everyone can verify that the sender wants to send her coins to the recipient, but a problem arises - how to prevent the inclusion of transactin Tx in the blockchain twice? Without such a protection an unscrupulous recipient may repeat Tx as long as the sender has enough coins at his balance, making it impossible to reuse the same address for more then 1 transaction. In particular the adversary can withdraw some coins from an exchange and repeat this transaction until there are no coins left on exchange (such attacks have already been…