DIP‐BLS: Dynamic Interaction Planning and BLS Aggregate Signatures for Low‐Latency Data Forwarding in Web 3.0

ABSTRACT The decentralized nature of Web 3.0 forms networks of autonomous, heterogeneous entities, including smart contracts and dapp modules, among others, which operate on untrusted infrastructure for interactive communication. In such environments, component churn is frequent, link conditions vary in real time, and the logical topology is perpetually in flux. Traditional routing paradigms that depend on static views or centralized controllers are ill‐suited for this setting, as they incur high overhead, exhibit poor adaptability, and introduce single points of failure. Meanwhile, ensuring end‐to‐end data authenticity and processing integrity across multiple relay hops remains a critical yet unresolved concern in the absence of trusted intermediaries. To tackle these intertwined challenges, we propose DIP‐BLS, a unified protocol that couples dynamic interaction planning with BLS aggregate signatures into a coherent data exchange framework. On the one hand, each node independently maintains a local interaction plan that tracks neighbor connectivity, latency, available bandwidth, and load through lightweight probes, enabling multi‐metric forwarding decisions without global coordination. On the other hand, every relay hop appends a BLS signature to the packet, and these signatures are progressively aggregated into a constant‐size credential, permitting efficient batch verification at the destination. Theoretical analysis and simulation experiments demonstrate that, compared with shortest‐path and flooding‐based methods, DIP‐BLS achieves marked improvements in data delivery ratio, average interaction latency, and system throughput, while providing strong tamper resistance and identity traceability at acceptable overhead—well‐suited to Web 3.0 distributed application scenarios that demand both high efficiency and robust security.