Composite α-μ Based DSRC Channel Model Using Large Data Set of RSSI Measurements

Abstract: Channel modeling is essential for design and performance evaluation of numerous protocols in vehicular networks. In this work, we study and provide results for largescale and small-scale modeling of communication channel in dense vehicular networks. We first propose an approach to remove the effect of fading on deterministic part of the large-scale model and verify its accuracy using a single transmitter-receiver scenario. Two-ray model is then utilized for path-loss characterization and its parameters are derived from the empirical data based on a newly proposed method. Afterward, we use {\alpha}-{\mu} distribution to model the fading behavior of vehicular networks for the first time, and validate its precision by Kolmogorov-Smirnov (K-S) goodness-of-fit test. To this end, the significantly better performance of utilizing {\alpha}-{\mu} distribution over the most adopted fading distribution in the vehicular channels literature, i.e. Nakagami-m, in terms of passing K-S test has been investigated and statistically verified in this paper. A large received signal strength indicator (RSSI) dataset from a measurement campaign is used to evaluate our claims. Moreover, the whole model is implemented in a reliable discrete event network simulator which is widely used in the academic and industrial research for network analysis, i.e. network simulator-3 (ns-3), to show the outcome of the proposed model in the presence of upper layer network protocols.