Multivariate $\alpha$-Stable Models in OFDM-Based IoT Networks with Interference From a Poisson Spatial Field of Interferers

The uncoordinated nature of IoT networks makes interference management a challenging problem. Motivated by NB-IoT and SCMA protocols, we study the interference statistics of a Poisson spatial field of IoT interferers exploiting OFDM. We show for a sufficiently large number of subcarriers that the interference statistics are well-approximated by a sub-Gaussian $\alpha$-stable random vector with a non-isotropic underlying Gaussian random vector. This result forms a basis to improve detection and decoding algorithms at the receiver.