Algorithm based on Byzantine agreement among decentralized agents (BADA)

Abstract

Distributed consensus requires the consent of more than half of the congress to produce irreversible results, and the performance of the consensus algorithm deteriorates with the increase in the number of nodes. This problem can be addressed by delegating the agreement to a few selected nodes. Since the selected nodes must comply with the Byzantine node ratio criteria required by the algorithm, the result selected by any decentralized node cannot be trusted. However, some trusted nodes monopolize the consensus node selection process, thereby breaking decentralization and causing a trilemma. Therefore, a consensus node selection algorithm is required that can construct a congress that can withstand Byzantine faults with the decentralized method. In this paper, an algorithm based on the Byzantine agreement among decentralized agents to facilitate agreement between decentralization nodes is proposed. It selects a group of random consensus nodes per block by applying the proposed proof of nonce algorithm. By controlling the percentage of Byzantine included in the selected nodes, it solves the trilemma when an arbitrary node selects the consensus nodes.

Physical Layer Security In Autonomous Driving Based Non-Competitive Distributed Consensus Scheme

“…Oh et al proposed a distributed consensus algorithm among Decentralized Agents (BADA) [13][14][15]. In the BADA distributed consensus algorithm, not all mode nodes participate in the consensus process, but consensus is reached through a committee composed of some nodes.…”

“…In this case, scalability is guaranteed by proposing a consensus protocol that provides message complexity for the exchange of consensus messages between distributed congress nodes. Accordingly, the BADA distributed consensus algorithm provides consensus time within several seconds and performance of thousands of TPS [13].…”

“…The BADA distributed consensus algorithm uses a nonce chain to verify the validity of a vehicle node [13,14]. The node generating the block must disclose the last nonce value and height value in its nonce chain.…”

“…However, research into such blockchain techniques is expected to have many di culties in sharing real-time tra c information. This is because the existing blockchain technology requires a lot of time before reaching a node consensus for block generation, making it di cult to share real-time tra c information [12], [13].…”

“…Blockchain research to provide safety for IoT services is active, but research into the real-time applicability of blockchain services is relatively insigni cant to that extent [13]. Our research basically presents a blockchain technology to implement real-time tra c information sharing for autonomous driving.…”