Proteus: An intelligent secret sharing orchestration framework for trusted integrated information services in the edge–cloud continuum

The rise of Integrated Information Services (IIS) within the Edge–Cloud Continuum demands stringent Quality of Service (QoS) alongside robust data security. While Secret Sharing (SS) provides strong theoretical guarantees, traditional static configurations suffer from severe computational bottlenecks and bandwidth contention in dynamic environments, disrupting real-time service continuity. To address this, we propose Proteus , an AI-driven adaptive orchestration framework for trusted IIS in the edge–cloud continuum. Proteus models parameter tuning as a Markov Decision Process (MDP) and employs a Proximal Policy Optimization (PPO) agent to dynamically co-optimize secret sharing thresholds and RaptorQ coding parameters, balancing computational overhead against communication redundancy. The Dumbo asynchronous consensus protocol is integrated to further minimize execution latency under unreliable network conditions. Experiments on a 100-node AWS testbed show that Proteus achieves up to a 15 × acceleration in reconstruction speed, cuts bandwidth consumption by up to 40%, and improves throughput by up to 31% compared to representative static baselines. These results suggest that Proteus serves as an effective orchestration framework for supporting Service Level Agreements (SLAs) in next-generation trusted information services.