Scheduling Optimization for Mixed-type Devices of IoT in Massive MIMO Systems with Spatio-Temporal Traffic

In this paper, we develop a framework for the analysis of massive multiple-input multiple-output (MIMO) systems where various types of devices with different configurations and requirements co-exist, taking into account both the spatial and temporal random traffic. A closed-form approximation of the spatial mean packet throughput which denotes the average number of packets that are successfully transmitted at any unit time slot and area is derived by using tools from stochastic geometry and queuing theory. Based on it, we investigate the optimal scheduling number for each type of device that maximizes the spatial mean packet throughput while meeting devices' delay demands. It is found that when the BS antenna number (M) goes to infinity, the BS should schedule all devices under its coverage, regardless of devices' variations. However, when M is finite, the BS should have a bias on scheduling devices with heavier arrival traffic and lower decoding threshold. On the basis of this, if the delay constraint of one device becomes stricter, the BS will allocate extra shares of the scheduling quantity to it, and this allocation acts more evidently with small-scale BS antennas.