This article introduces a new approach to address the spectrum scarcity challenge in 6G networks by implementing the enhanced licensed shared access (ELSA) framework. Our proposed auction mechanism aims to ensure fairness in spectrum allocation to mobile network operators (MNOs) through a novel weighted auction called the fair Vickery-Clarke-Groves (FVCG) mechanism. Through comparison with traditional methods, the study demonstrates that the proposed auction method improves fairness significantly. We suggest using spectrum sensing and integrating UAV-based networks to enhance efficiency of the LSA system. This research employs two methods to solve the problem. We first propose a novel greedy algorithm, named market share based weighted greedy algorithm (MSWGA) to achieve better fairness compared to the traditional auction methods and as the second approach, we exploit deep reinforcement learning (DRL) algorithms, to optimize the auction policy and demonstrate its superiority over other methods. Simulation results show that the deep deterministic policy gradient (DDPG) method performs superior to soft actor critic (SAC), MSWGA, and greedy methods. Moreover, a significant improvement is observed in fairness index compared to the traditional greedy auction methods. This improvement is as high as about 27% and 35% when deploying the MSWGA and DDPG methods, respectively.

翻译：本文提出了一种新方法，通过实施增强型授权共享接入（ELSA）框架来解决6G网络中的频谱稀缺挑战。我们提出的拍卖机制旨在通过一种名为公平维克里-克拉克-格罗夫斯（FVCG）的新型加权拍卖，确保移动网络运营商（MNO）在频谱分配中的公平性。通过与传统方法比较，研究表明所提出的拍卖方法显著提高了公平性。我们建议利用频谱感知并集成基于无人机的网络来提升LSA系统的效率。本研究采用两种方法解决该问题：首先提出一种新颖的贪心算法——基于市场份额的加权贪心算法（MSWGA），以实现优于传统拍卖方法的公平性；其次，利用深度强化学习（DRL）算法优化拍卖策略，并证明其优于其他方法。仿真结果表明，深度确定性策略梯度（DDPG）方法的表现优于软演员-评论家（SAC）、MSWGA和贪心方法。此外，与传统的贪心拍卖方法相比，公平性指标显著提升，采用MSWGA和DDPG方法时，提升幅度分别高达约27%和35%。