This paper extensively examines the impact of the numerology schemes on a sliced TDD radio access network. The slices are multiplexed over both the sub-6 GHz and mmWave bands. The incorporation of numerology schemes into network slicing, the impact of which has yet to be explored, will be highly instrumental in realizing the future 5G and 6G networks. Thus, this work demonstrates the enhanced capabilities of sliced networks utilizing numerology. To that end, two optimization problems are formulated, one to maximize the downlink average spectral efficiency of a throughput-sensitive slice, and the other to minimize the uplink transmission power of the users of the latency-sensitive slice. This work optimizes key parameters such as duplex ratio, numerology scheme and optimally allocates power and bandwidth under numerology-enabled slicing. The comparative analyses highlight that for high-throughput applications, lower numerology schemes are significantly more spectrally efficient and can often achieve throughputs characteristic of higher schemes. On the other hand, the higher schemes are the best choice for the minimization of user transmission power and latency, which are crucial in energy-efficient communications and edge computing applications, respectively.

This work proposes a dynamic numerology scheme assignment framework to provision mobile-edge computing (MEC) for massive Internet of Things (IoT) networks via unmanned aerial vehicles (UAVs). IoT devices (IoTDs) usually lack computational power; thus, they offload their computational tasks to MEC servers. To enhance their battery lives, an optimal assignment of a numerology scheme to each IoTD is imperative; it also enhances the system spectral efficiency. In this work, we bring MEC services closer to a massive IoT network by deploying a UAV-MEC and allocate communication resources of the sub-6-GHz band dependent on the numerology schemes to each IoTD. We formulate a multiobjective optimization problem (MOOP) with two countering objectives to maximize the uplink spectral efficiency while minimizing the IoTDs’ energy consumption. We solve a series of sequential subproblems, which are convex approximations of the MOOP and propose a novel algorithm to allocate computational resources, assign numerology schemes, and communication resources for each of the IoTDs. Our extensive simulation results validate the proposed aims herein.