Weibull distributed recurrent neural ergodic skewed certificateless signcryption for data protection in cloud computing environment

Certificate-based cryptography dominates cloud platforms but introduces validation latency, revocation overhead, and residual escrow risk that threatens tenant confidentiality. Recent certificateless signcryption removes external certificates yet still requires multi-stage key negotiation, leaving setup slow and false-accept rates high under heavy-tail traffic. We propose a Weibull-Distributed Recurrent-Neural Ergodic-Skewed Certificateless Signcryption (DRESCS) pipeline that generates session keys in constant time, refreshes signing secrets every 128 cycles to eliminate escrow, and applies a Weibull activation layer to suppress out-of-distribution anomalies. The design is evaluated using 1,000 Monte Carlo simulations with realistic multi-tenant traces. DRESCS reduces key-setup latency by 28 percent, lowers false-accept rate to 0.8 percent, and achieves 163 signcrypt operations per millisecond, outperforming the strongest published baseline by 34 percent. Performance degrades by less than one percent under five percent injected noise. These results demonstrate fast, robust, and escrow-free protection without certificate management, making DRESCS suitable for latency-sensitive edge–cloud deployments and acceleration.