Time-Dependent Propagation Analysis and Modeling of LPWAN Technologies

Contemporary low-power wide area network (LPWAN) technologies have been introduced as connectivity enablers with low complexity, extended communication range, and excellent signal penetration. On the other hand, they suffer from a substantial delay and low packet-delivery guarantees. As a result, numerous novel applications entering the Internet of things (IoT) market suffer from insufficient performance. To mitigate this issue, further optimization and adaptation of the LPWAN technologies to the needs of these new applications requires an indepth understanding of the propagation environment dynamics. Motivated by that, in this paper, we thoroughly investigate timedependent statistical characteristics of the reference signal receive power (RSRP) dynamics of Narrowband IoT (NB-IoT) technology. We demonstrate that even for a stationary user equipment, RSRP is subject to drastic variations that are characterized by exponentially decaying autocorrelation function. We then demonstrate that first- and second-order statistical properties of the RSRP dynamics can be closely captured using a doublystochastic Markov model that retains the tractability of the conventional Markov models. The reported model is expected to serve as a building block for analytical and simulation-based system-level studies and optimization of LPWAN technologies.