30-05-2023 дата публикации
Номер: CN116193546A
Принадлежит:
The invention relates to a dynamic generalized user NOMA grouping CCHN-MEC network unloading decision optimization method, which comprises the following steps: firstly, under a given calculation unloading ratio, deriving to obtain an optimal solution of a local energy consumption minimization problem, i.e., local optimal CPU frequency allocation of a secondary user SU, secondly, solving the unloading energy consumption minimization problem through a convex optimization tool, and finally, solving the optimal CPU frequency allocation of the secondary user SU. The transmitting power and the transmission time of the CPU frequency SU which is calculated and allocated for each SU task in each NOMA group are obtained; finally, user calculation unloading ratio allocation of each time slot is learned according to a deep reinforcement learning algorithm based on SAC, and an optimal unloading decision is obtained. According to the invention, the energy consumption of the system can be greatly saved ...
Подробнее