A Hybrid TDMA-NOMA Scheme for Wireless Content Caching Networks

Non-orthogonal multiple access (NOMA) has been employed in wireless content caching networks to deliver requested files to users efficiently. However, its performance is constrained by the requirement that cached files must wait for uncached files to be retrieved from the cloud before delivery can commence. To address this limitation, we propose a hybrid time orthogonal multiple access (TDMA) - NOMA scheme that dynamically forms NOMA groups in a time-division manner, based on the availability of files at the base station, while simultaneously fetching uncached files from the cloud. We optimize the caching decision, file scheduling (user grouping), transmit power allocation, and time allocation to minimize average system latency. This is achieved through a divide-and-conquer approach utilizing dynamic programming, one-dimensional search, and successive convex approximation. Simulation results demonstrate that the proposed hybrid TDMA-NOMA scheme outperforms the existing hybrid NOMA, NOMA and orthogonal multiple access schemes. The proposed scheme is particularly effective in challenging scenarios characterized by small cache capacity, a large number of users, and large file sizes.