Novel Downlink Multiuser Resource-Allocation Scheme for Providing Layer-Encoded Multimedia Streams Using Massive MIMO Transmissions

Mobile video streaming is an intriguing application for next-generation networks. Wearing the goggles that render two-eye videos, users can enjoy the interactive multimedia experience. Providing high-quality video streams to multiple mobile devices in specific areas will become popular in future cinemas, theme parks, and museums. To ensure quality wireless coverage for a good streaming experience, the next-generation wireless technology (e.g., 5G/6G) employing massive MIMO schemes is a promising solution. Massive MIMO transmissions can improve bandwidth utilization while maintaining acceptable system complexity through numerous transceiving antennae. To incorporate massive MIMO transmissions with mobile video streaming, an innovative cross-layer scheme is needed to flexibly and efficiently manage the antenna array for serving multiple user devices. This allocation mechanism must have low computational complexity and operate stably to prevent demand fluctuations from affecting the quality of service experienced by other users. In this work, we introduce a new problem of provisioning layer-encoded streams to mobile devices (e.g., VR goggles) by allocating antennae in the base stations’ massive MIMO arrays. Given each user’s bitrate demand, the available antennae of each femtocell, and the channel characteristics, the system allocates transmitting antennae to maximize the total system utility. Our theoretical analysis shows that this allocation problem is NP-hard but our proposed scheme provides bounded performance with polynomial-time complexity. We also discuss and justify the stability of our proposed new allocation mechanism. Simulations demonstrate that our scheme outperforms simple heuristic methods. To the best of our knowledge, this is the first attempt to tackle antenna allocation for mobile user devices in immersive video streaming using massive MIMO schemes.