Cardano is labeled a decentralized open-source platform for smart contract applications. Like many other projects, its blockchain protocol is structured into two layers. A settlement layer handling all the transactions with the Cardano asset (its cryptocurrency) and a computational layer that hosts smart contracts. At the same time, Cardano’s network knows three different types of nodes: mCore nodes participate and handle Cardano’s blockchain governance; relay nodes are the ones that send data between mCore nodes and the public internet. And there are also edge nodes, which create and process the transactions. With this layered structure and specialized network nodes carrying out different tasks, the project aims to ensure high scalability while maintaining security.

Cardano’s main goal is to offer new ground-breaking features, but on a scalable public blockchain that is very secure. According to proponents, Cardano should one day be one of the most secure decentralized proof-of-stake blockchains. What is said to be special about Cardano’s blockchain, is its consensus protocol, which has been developed in-house. It goes by the name of Ouroboros and is deemed by its developers “the first probably secure proof-of-stake protocol”. The main reason for this, people behind Cardano argue, is the fact that Cardano is being developed by top-notch researchers and on a peer-reviewed basis. Unlike other cryptocurrency projects, where practice preceded formal theory, developers behind Cardano aim to construct its blockchain by letting formal theory precede practice, and prove each security claim to be mathematically correct before implementing it.  At the same time, Cardano applies a layered infrastructure (as described above) from the get-go. Because not all of Cardano’s eras have been implemented and the protocol is still in its early stages, it is quite hard to put this claim to the test. But notably, Cardano’s approach has been criticized by none other than Vitalik Buterin.

The way Ouroboros works is that time is divided up into epochs and slots. The epochs represent overarching time frames, while the slots are 20-second increments within each epoch. Within a slot, slot readers get randomly chosen and they are responsible for choosing blocks. The more tokens a person holds, the better the opportunity they get chosen as a slot leader. Ultimately, slot readers can be likened to validators. A slot leader generates new blocks on the Cardano blockchain and verifies the transaction. Anyone holding Cardano’s token can become a stakeholder and can also become a slot leader. It’s those mCore nodes already mentioned that can get picked as slot readers. From a bird’s eye perspective, this is the way consensus is achieved in Cardano.

With Cardano’s consensus protocol Ouroboros there is not only one, but two types of blocks added to the blockchain. There are genesis blocks, and these include a list of all eligible slot readers of one epoch. A genesis block also consists of a bunch of main blocks. These are the second type of blocks and they contain all transaction information as well as potential proposals for software updates. With every epoch coming to an end, it is upon the current slot readers to elect the future slot readers of the next epoch. Every time they perform such an election, slot readers stick to a mechanism similar to a coin tossing act. As indicated by Cardano developers this helps increase the randomness by which new slot readers are chosen.

And once the implementation of Ouroboros is finalized, it is expected to enable near-instant finality. What is interesting is the fact that, although Cardano is a proof of stake blockchain, block validation resembles that of proof of work blockchains. The Ouroboros consensus algorithm uses probabilistic finality, making reorganization hypothetically possible due to the blocks being added canonically as opposed to validators finalizing the blocks as they are produced. This is similar to how the Bitcoin blockchain is being built block by block.