IMPLEMENTING DISTRIBUTED GENETIC ALGORITHM WORKFLOW WITH CHANNELS AND MESSAGE PROCESSING NODES

The implementation of the genetic algorithm using the architectural design pattern "Pipes and Filters" solves two main problems – efficiency and ease of modification. The first is achieved due to asynchronous communication and the ability to scale individual phases by adding new nodes of the selected type. The second is ensured by a low level of connectivity of the system components among themselves, namely, the nodes should know only the format and protocol, based on which they should form messages with the results of their work. Modern languages and programming environments provide a large set of tools for implementing systems of this type, completely abstracting from interaction with the operating system kernel. The most leading programming language for implementing distributed computing is Golang.