Transaction validation in blockchain networks is based on the formation of P2P groups. Such principle of operation was already discussed in our previous article “Introducing Proof-of-Unity” , which presented Prometeo’s new consensus mechanism.

Most, if not all, blockchain networks currently rely on mathematical solutions for P2P validations, and all nodes traditionally are assigned randomly or based on predetermined criteria, to form said P2P groups. However, such an approach leaves networks vulnerable to external single attacks, which means that one needs to only successfully attack a single node, to gain access to the whole network, eventually jeopardizing the integrity and security of the blockchain. Such exposed vulnerabilities may lead not only to leakages of data and money but also considerably slow down the efficiency of the system.

Prometeo Chain System’s team already thought ahead about this issue, and we are pleased to announce our new latest project: an AI technology that will be responsible for the optimization of P2P group validations, a solution that was never done before.

By utilizing deep reinforcement learning algorithms, our AI agent selects the best nodes for the formation of P2P group validators by evaluating a multitude of factors, some of which are: node lifetime, transaction history, reputation score, reputation history, security, processing capacity, and many more. This intelligent selection process (which will grow refined each time AI agent is used) will significantly reduce the amount of failed P2P group validations and enable faster and more efficient validation, unlocking significant advantages for participating nodes.

Moreover, the AI agent will be designed to be inclusive in that, all nodes, be they supported by computers or smartphones, or other devices, won’t be excluded from forming newly optimized P2P groups. In that way, no network stakeholder will be ignored on the basis of the device type used in the decisions of our AI agent.

This dynamic approach also has the side effect of reducing the risks posed by single attacks by adding an external agent that will also oversee any kind of malfunction or malicious attacks and won’t allow the compromised transaction to go through. In the future, with Prometeo's AI, P2P networks can confidently operate with heightened protection against vulnerabilities and ensure the integrity of their transactions.

Now, to make full use of our AI’s capabilities, nodes will need to grant the agent access to their data. The security of the data extracted by AI and its unavailability to any third parties will be ensured by signing a contract with the developer of AI.

Since the agent is developed exclusively by our team, this AI will also belong to Prometeo Chain System.

Every network node possesses unique characteristics depending on its activity and the type of network it operates within. Some potential features were already mentioned before (like node lifetime, transaction history, reputation score, etc), but these are not nearly all factors that our AI takes into account. Below you can see a table with some of the node characteristics based on which AI determines the best nodes:

By considering these diverse sets of characteristics, our AI technology ensures an intelligent and comprehensive selection of nodes, leading to enhanced P2P validation efficiency.

We are thrilled to be at the forefront of this innovative AI approach and learn more about the opportunities given by the deep reinforcement learning methodology.

Stay tuned for more updates as we continue to push the boundaries of AI applications in P2P optimization!