AlephBFT Consensus

The code behind the Aleph Zero blockchain is AlephBFT, which is a peer-reviewed consensus protocol that was presented at the Advances in Financial Technologies (AFT) 2019 conference in Zurich.

Aleph Zero takes advantage of a heavily modified variation of the Proof of Stake (PoS) consensus methodology in conjunction with Directed Acyclic Graphs (DAG). At its core, Aleph Zero employs a PoS consensus algorithm in which a rotating committee of validators decides on the truthfulness of the state of the blockchain. This makes it similar enough to other blockchains employing PoS methodologies. So why does Aleph Zero use a DAG?

DAG acts as an intermediary in the creation of the Aleph Zero blockchain. These data structures are primarily associated with the world of mathematics and computer science but have recently made waves in the blockchain industry. From a technical standpoint, using DAGs allows for faster transaction speeds and higher throughput and hence significantly lower transaction fees. Mining is a costly process, resulting in Proof of Work (PoW) algorithms generating high transfer costs. This is one of the advantages of DAGs. The low cost of the technology makes it feasible for mass adoption.

On the Order of Things

One of the biggest obstacles towards scaling blockchains is effectively ordering the transactions conducted on the network. The process of properly ordering transactions is essential to prevent double-spending and other malevolent practices. By taking advantage of DAGs, we can eliminate the threat of malicious block producers and create truly leaderless protocols. The chain itself forms afterward. This results in the creation of two structures simultaneously. The DAG is an auxiliary structure that collects information on the order of transactions. The chain itself is constructed from the collected data. Validators propose blocks for the chain, always building on the highest correct blocks known to them. The DAG impacts the construction of the chain, yet is not the chain itself with the blocks finalized using DAG as an auxiliary structure in the parallel process.

One of the challenges facing DAG solutions is their purported lack of decentralization. In fact, this problem doesn’t concern DAG-based protocols any more than any other PoS systems; it just happened so that the early DAG-based solutions found their first application in a permissioned setting.

The Aleph Zero development team has laid this claim to rest by ensuring a new standard of decentralization for DAGs. How did we manage to achieve this? To guarantee our protocol remains decentralized, we have adopted a mechanism that appoints a rotating committee of random members to decide on the future of the network. Conservative figures for this committee start at 128 nodes, but we will be aiming for much higher numbers to spread out the network as much as possible.

AlephBFT and its Properties

The second crucial ingredient of our consensus protocol is the Byzantine Fault Tolerance (BFT) architecture we employ. Through the use of BFT, we can ensure that the communication between nodes remains effective and transparent despite the presence of malicious nodes. BFT consensus models suppose that less than ⅓ of the nodes within a network are acting in bad faith. If we assume ⅔ of the users are honest, a BFT blockchain can operate securely while ensuring an immutable and proven history of transactions.

One of the criticisms directed towards BFT solutions is their perceived ineffectiveness beyond permissioned settings. We have faced this challenge through the introduction of rotating committees. Additional security is guaranteed by employing the element of asynchronicity. This means that the network will work smoothly and retain data correctness even if some parts of the network are down. Implementing asynchronous properties to BFT is a step forward to ensure maximum security. This benefits enterprise users as well as open-source projects like Aleph Zero.

Industry Adoption of BFT Solutions

The blockchain space is seeing a wider adoption of BFT solutions. This includes Polkadot’s very own GRANDPA/BABE solution as well as the Casper Highway Protocol.

• In Polkadot’s case, their BFT proposal works as a partially synchronous network model when ⅔ of the nodes are honest and can manage ⅕ Byzantine nodes in an asynchronous setting. They can expedite the finalization process by reaching a consensus on chains and not on individual blocks.

• The Casper Network prides itself on higher levels of finality and flexibility than other BFT protocols currently on the market. One aspect of their unique take on BFT is incentivizing honest behavior by offering higher finality when more nodes behave honestly. To maintain flexibility, individual nodes can decide on their finality threshold.

This means that nodes can play different roles in the ecosystem. A node can require higher finality for transactions that are larger, while less essential transactions can have a lower threshold.

Defending the Blockchain

Malicious nodes and users operating in bad faith will always be present, no matter how secure a consensus protocol is. In this sense, the blockchain space mirrors the real world with the caveat that blockchain technology offers unmatched security solutions. It is important to discuss the “nuclear options” a blockchain has at its disposal when dealing with malicious nodes or nodes that draw suspicion of acting against the interest of honest users. Aleph Zero distinguishes itself in this respect by committing itself to the idea of freezing funds rather than slashing them in the case of suspected malpractice.

As with most PoS protocols, there are certain dangers we must consider that may arise from users with less than pure intentions. Most PoS protocols employ slashing as a potential punishment for corrupt behavior. Slashing involves confiscating the funds allocated by an individual to take part in the staking mechanism. This is a shortsighted practice that can result in unnecessary accusations and financial loss. Often, the accused are the victims of errors in the protocol and receive punishment for occurrences beyond their control.

Our official stance on this matter is that we don't have and are not planning to introduce an automatic slashing mechanism. To combat malicious actors, Aleph Zero’s on-chain governance mechanism (which the team aims to implement as soon as possible) might employ a freezing tool instead. Freezing funds would involve the temporary suspension of capital staked by the suspected user. Firstly, an investigation will take place that will determine whether an internal error occurred or whether the user had malicious intentions. This case-by-case approach is a significantly fairer approach to the matter than indiscriminate slashing. After the investigation is concluded, we will enforce fund slashing or release them back to the owner (in case the suspect occurs to be innocent and the fault is attributed to a bug in the code, for example). Initially, the decision will be up to us, eventually, decentralized governance will decide upon it.