TAXONOMY AND ARCHITECTURAL CLASSIFICATION OF BLOCKCHAIN TECHNOLOGIES: OPPORTUNITIES AND FUTURE APPLICATIONS

Abstract

Modern centralized computing services do not provide a sufficient level of decentralization, transparency, and scalability, which limits their use in open ecosystems. There is a growing need for a well-founded selection of a blockchain architecture for the development of an efficient computing power marketplace that combines security, performance, and economic viability. The subject of the present analytical study is the architectural characteristics and evaluation metrics of blockchain technologies for building decentralized platforms for computational resource exchange.The goal of the study is to systematize knowledge of blockchain architecture types, consensus algorithms, and performance metrics to justify the technological basis for the creation of a custom computing power marketplace.The tasks addressed in the paper include: classification of blockchain architectures by openness, consensus type, and scalability approach; comparative analysis of technical, functional, and economic metrics of leading platforms (Ethereum, Polkadot, Hyperledger Fabric, IOTA, Solana); review of existing marketplaces such as Akash, Golem, and iExec to identify their strengths and limitations; and formulation of requirements for a custom solution.As a result of the analytical review, a systematic typology of blockchain solutions was developed and architectural comparison metrics were generalized with respect to the blockchain trilemma (decentralization – scalability – security).A comparative analysis of five modern platforms (Ethereum, Hyperledger, Polkadot, Solana, IOTA) was conducted in terms of their applicability for a computing power marketplace.It was found that DAG-oriented architectures (e.g., IOTA) offer higher scalability potential but require improvements in security mechanisms. BFT-based platforms demonstrate high throughput but are limited in terms of open access.Principles for constructing a hybrid architecture based on edge computing were proposed, enabling efficient load balancing, reduced latency, and greater autonomy among participants in a distributed computational environment.Conclusions: The application of a typified approach to selecting blockchain architecture enables the avoidance of common limitations of existing systems and ensures high adaptability, efficiency, and transparency in the new computing power marketplace.