Fog Computing (FC) is a paradigm that offers centralized cloud computing services to the proximity of a device, i.e., at the edge for fast, secure, efficient, and reliable processing. Although FC offers many benefits, but it is not ideally suitable for various mission-critical applications such as healthcare, autonomous vehicle, connected cars, and unmanned aerial vehicles. To overcome the aforementioned issue, this paper integrates the concept of Device-to-Device (D2D) communication with the traditional FC. D2D offers faster processing due to the local processing of data at the device itself. But, D2D communication is potentially affected by interferences (from the cellular user and other devices) and noises (additive-white Gaussian noise). This can be mitigated if the Orthogonal Frequency Division Multiple Access (OFDMA) is replaced with Non-Orthogonal Multiple Access (NOMA) scheme. The Successive Interference Cancellation (SIC) technique of NOMA minimizes the interference effect. To improve the secrecy capacity of the network and spectral efficiency, we used a coalition game theory. Simulation results scrutinize the superiority of the proposed NOMA-based D2D-aided FC system over sum rate, secrecy capacity, and processing delay parameters.

雾计算（FC）是一种范式，它向设备的附近（即在边缘）提供集中的云计算服务，以进行快速，安全，高效和可靠的处理。尽管FC提供了许多好处，但是它并不理想地适合于各种关键任务应用，例如医疗保健，自动驾驶汽车，联网汽车和无人驾驶飞机。为了克服上述问题，本文将设备到设备（D2D）通信的概念与传统FC进行了集成。由于设备本身对数据进行本地处理，因此D2D提供了更快的处理速度。但是，D2D通信可能受到干扰（来自蜂窝用户和其他设备的干扰）和噪声（加性白高斯噪声）的影响。如果将正交频分多址（OFDMA）替换为非正交多址（NOMA）方案，则可以缓解这种情况。NOMA的连续干扰消除（SIC）技术使干扰影响最小化。为了提高网络的保密能力和频谱效率，我们使用了联盟博弈理论。仿真结果仔细研究了所提出的基于NOMA的D2D辅助FC系统在总速率，保密容量和处理延迟参数方面的优越性。