Hydra Radio Access Network (Hydra-RAN): Multi-Functional Communications and Sensing Networks, Adaptive Time-To-Trigger Optimization

Seamless mobility management is a critical challenge in next-generation multi-functional (NG-MF) wireless networks, where traditional handover schemes with static Time-To-Trigger (TTT) parameters prove inadequate in dynamic environments. These inflexible settings often result in elevated handover latency, connection failures, and ping-pong effects, ultimately degrading Quality of Service (QoS). To address this, we propose a novel adaptive and autonomous TTT optimization framework for the Hydra-RAN architecture, a multi-functional network that integrates sensing and communication capabilities. Our framework leverages real-time sensor data and channel state information (CSI), processed by a Deep Reinforcement Learning (DRL) agent built on a Sparse Multi-Task Learning (SMTL) model. This SMTL-DRL engine dynamically predicts optimal TTT values by selecting from a pre-defined codebook, enabling the network to intelligently balance handover latency and reliability. It is particularly suited to the complex demands of NG-MF wireless networks since it dynamically adjusts TTT parameters to minimize disruptions and maintain high quality of service. Simulations demonstrate that our proposed solution significantly outperforms conventional static TTT approaches, achieving substantial reductions in handover latency and ping-pong rate while ensuring robust QoS across diverse user mobility scenarios.