In recent years, blockchain technology, which enables transactions to be distributed across multiple computers and managed in an immutable and secure manner, has garnered significant attention. Within blockchain networks, consensus mechanisms ensure the consistent sharing of ledger information when new blocks are added. In hybrid blockchains, typically employed by a limited number of organizations, the Practical Byzantine Fault Tolerance (PBFT) protocol is widely used. PBFT is designed to tolerate Byzantine nodes-nodes that may be compromised or malfunctioning-by achieving consensus as long as fewer than one-third of the total nodes are Byzantine. However, PBFT relies on the assumption that at least two-thirds of the nodes behave correctly, making consensus challenging when the number of malicious nodes exceeds this threshold. Previous research has explored the use of clustering to enhance throughput, but these methods are static and unsuited to dynamic environments. Moreover, clustering techniques aimed at bolstering Byzantine resistance remain underexplored. This paper presents a novel method for constructing clusters within a blockchain network to resist Byzantine nodes. By employing clustering, we estimate the locations of potential attackers, thereby enhancing the system's resilience to Byzantine faults.

The increasing adoption of blockchain technology has underscored the need for cross-chain systems that enable seamless communication among multiple BC networks. Achieving cross-chain interoperability, which ensures secure and efficient data storage and transfer across BCs, remains a critical technical challenge. Among existing solutions, the Inter-Blockchain Communication (IBC) protocol within the Cosmos ecosystem is a prominent framework that facilitates cross-chain communication using light clients and continuous monitoring. However, IBC faces limitations due to its reliance on the consensus algorithms and block generation intervals of participating blockchains. Cross-chain transactions are processed sequentially, requiring approval at each stage, which reduces efficiency. Furthermore, the increasing number of relayers introduces scalability and operational challenges. To address these issues, this study proposes a novel framework called Relayer Aggregation (RA), which aims to enhance cross-chain communication by employing a chainless multi-layer consensus mechanism. RA enables parallel transaction processing to improve performance and scalability. Experimental nodes were developed, and comparative performance evaluations of RA and IBC were conducted to validate the proposed approach. The results demonstrate that RA significantly reduces the number of required relayer nodes and enhances processing efficiency through parallelization. By overcoming the sequential processing limitations of IBC, RA offers a scalable and efficient solution for cross-chain interoperability. This study contributes to advancing blockchain ecosystems by addressing key bottlenecks in cross-chain systems and providing a foundation for future optimization in distributed environments.