Abstract
Web 3.0 is recognized as a pioneering paradigm that empowers users to securely oversee data without reliance on a centralized authority. Blockchains, as a core technology to realize Web 3.0, can facilitate decentralized and transparent data management. Nevertheless, the evolution of blockchain-enabled Web 3.0 is still in its nascent phase, grappling with challenges such as ensuring efficiency and reliability to enhance block propagation performance. In this paper, we design a Graph Attention Network (GAT)-based reliable block propagation optimization framework for blockchain-enabled Web 3.0. We first innovatively apply a data-freshness metric called age of block (AoB) to measure block propagation efficiency in public blockchains. To achieve the reliability of block propagation, we introduce a reputation mechanism based on the subjective logic model, including the local and recommended opinions to calculate the miner reputation value. Moreover, considering that the GAT possesses the excellent ability to process graph-structured data, we utilize the GAT with reinforcement learning to obtain the optimal block propagation trajectory. Numerical results demonstrate that the proposed scheme exhibits the most outstanding block propagation efficiency and reliability compared with traditional routing mechanisms.
The Challenge of Decentralized Propagation
In a decentralized Web 3.0 environment, blockchains rely on miners to broadcast new blocks to the rest of the network. Currently, most blockchains use "Gossip" protocols—essentially a random broadcast. This randomness creates two major headaches:
- Efficiency Gaps: If a block takes too long to reach everyone, the network can "fork" (split into two versions), leading to failed transactions and wasted energy.
- Reliability Risks: Random broadcasting doesn't account for malicious actors. Malicious miners can delay blocks or launch Sybil attacks, eroding the trust essential for Web 3.0.
The authors tackle the challenge of finding an "optimal trajectory"—the best sequence of miners to pass a block through—to ensure it arrives quickly and only via trustworthy nodes.
A Three-Layered Optimization Framework
The proposed solution is a multi-layered framework (Physical, Data, and Network) that utilizes several advanced techniques to ensure blocks are propagated efficiently and securely.
Key Components:
- Age of Block (AoB): Borrowing from the "Age of Information" (AoI) concept in communication theory, the authors created AoB. Unlike simple latency, AoB measures the "freshness" of a block by accounting for the time elapsed from a request to successful receipt, including waiting times in queues.
- Subjective Logic Reputation: To filter out bad actors, they implemented a reputation mechanism. It uses "Subjective Logic" to combine a miner’s direct experience (local opinions) with feedback from other peers (recommended opinions). Crucially, it weights negative interactions more heavily and prioritizes recent "fresh" interactions.
- GAT-based Reinforcement Learning: The miner network is treated as a graph. A Graph Attention Network (GAT) extracts structural relationships between miners, while a Reinforcement Learning (RL) agent uses these features to select the next miner in the propagation path.
Technical Implementation: GAT and RL
The core of the optimization lies in the GAT model's architecture. The Encoder uses multi-head attention to process the graph-structured data of the miner network, extracting deep structural features. The Decoder then functions as an RL agent, performing step-by-step selection of the next miner in the sequence to minimize total AoB while maintaining high reputation standards.
Key Results and Insights
The proposed GAT-based model was compared against standard "Greedy" and "Gossip" mechanisms in simulated networks of up to 49 miners. The results highlight significant improvements:
- Efficiency: The overall AoB was 2.4% to 22.6% lower than traditional methods, reducing the probability of network forks.
- Reliability: The total reputation value of the chosen miners was 7.4% to 34.77% higher, effectively routing blocks around low-reputation or malicious nodes.
- Global Optimization: Unlike "Greedy" algorithms that only look at the nearest neighbor, the GAT model considers the entire graph structure, finding paths that are faster and more reliable overall.
Future Directions
While the results are promising, the authors acknowledge the need for further scalability to handle the thousands of nodes found in real-world public blockchains. Future work will explore "sparse" GAT variants and adapt the framework for Proof of Stake (PoS) and other emerging consensus mechanisms.
Summary updated