Optically enhanced distributed swarm-optimized data stream framework for real-time computing

The growing complexity and speed of real-time data streams in IoT and distributed computing environments demand optimization frameworks that combine computational efficiency with adaptability to dynamic network conditions. To overcome the latency and scalability bottlenecks of traditional centralized or statically scheduled systems, this study proposes an Optically Enhanced Distributed Swarm-Optimized Data Stream Framework (O-DSO-Stream) that integrates decentralized swarm intelligence with optical computing acceleration and photonic interconnects. The framework incorporates an adaptive particle cooperation mechanism and a multi-objective optimization model to jointly minimize latency and energy consumption while preserving prediction accuracy. Optical computing modules are employed for parallel feature aggregation, and optical inter-node communication channels reduce synchronization delays and data transmission overhead. Experiments conducted on Kafka-based IoT streams and StreamDM benchmarks show that O-DSO-Stream achieves 22.8% faster convergence, 15.3% higher throughput, and 11.8% lower energy consumption compared with the Reinforcement-Stream baseline (p ⪅0.01, n=10). The integration of photonic hardware significantly improves system responsiveness and energy efficiency in heterogeneous clusters. These results demonstrate that O-DSO-Stream provides a scalable, low-latency, and resource-efficient solution for real-time distributed data optimization, with promising applications in optical sensor monitoring, photonic computing networks, and adaptive analytics systems.