Semantic-Empowered Status Update System in Satellite-Integrated Internet Under UoI Minimization

Considering the demands of conveying accurate information and goal-relevant data at minimal cost intelligently for the upcoming satellite-integrated Internet, we propose a new semantic-empowered metric called utility loss of information (UoI), which can simultaneously quantify the value loss due to the duration and severity of mismatch transceivers, and evaluate the energy loss of unnecessary data generation and transmissions. First, we develop a joint UoI-optimal problem to efficiently generate and transmit data for status updates with a multi-state Markov source, aiming to achieve an optimal tradeoff between value and energy loss of information. Then, we reformulate the above problem to a constrained Markov decision process (CMDP) due to the limited power and prove it has a threshold structure. Further, we propose a simplified relative value iteration (SRVI) algorithm to calculate the thresholds. Finally, by leveraging the derived generation and transmission ratio, we utilize the bisection search combined with our SRVI to find the CMDP UoI-optimal policy. Simulation results validate the optimal tradeoff achieved by our UoI-optimal policy in terms of timeliness, reliability, and energy-efficiency, as well as the superior performance compared to several state-of-the-art semantic-aware policies.