Article Preview
Top1. Introduction
Vehicular propagation channel model plays an important role in the performance evaluation of Vehicle-to-Vehicle (V2V) communication systems (Molisch, 2009) (Mir, 2014). The future advancements of such systems rely on their ability to perform well in different environments such as highway, urban, rural, and suburban. The key challenge is to model an accurate and efficient vehicular channel that can take into account random and time-varying characteristics of different environments. A precise radio channel model leads the way toward better understanding and therefore designing of the vehicular communication systems (Al-Bado, 2012).
Propagation channel modeling is a well-established research field, and researchers have presented several approaches in the literature. The work on vehicular channel modeling can be categories into three classes. A stochastic or analytical approach models the radio propagation characteristics by mean of statistical techniques such as channel estimation and equalization. Due to its low computation complexity, stochastic models often find their application in large-scale system-level simulations. However, the analytical approaches have limitations mainly because of their inability to model the environment-specific channel characteristics accurately. The deterministic or empirical models employs techniques such as 3D ray tracing to compute propagation characteristics of a specific environment. The environment-specific channel modeling approaches achieve higher accuracy at the expense of higher computational complexity. Finally, the geometry-based or semi-empirical channel modeling approach treads through the crossroad of analytical and deterministic channel modeling techniques. The idea is to use analytical and empirical models to determine different propagation mechanisms such as large-scale and small-scale fading. Moreover, the geometry-based approaches provide a good trade-off between computational complexity and accuracy.
The geometry-based channel models are capable of representing the propagation environment more accurately. To this end, these modeling techniques rely on a variety of information regarding surrounding of the vehicles. This information includes building geometry data obtained from both commercial and publicly available sources (Boban, 2014) (Sommer, 2011) (Nuckelt, 2013). Previously, satellite images have also been used to model the V2V channel environment accurately (He, 2012) (Wang, 2012). The environment-specific channel modeling techniques utilize realistic vehicular mobility traces to perform channel characterization at a citywide scale. In (Boban, 2014) Mobon et al. proposed an efficient geometry-based propagation model (GEMV2). GEMV2 uses building geometry and vehicle outlines to differentiate three types of links, i.e., line-of-sight (LOS), non-line-of-sight links obstructed by buildings (NLOSb) and non-line-of-sight links obstructed by vehicles (NLOSv). Since each link type possesses different channel characteristics, the propagation mechanisms such as reflection, diffraction and scattering are calculated using both deterministic and stochastic modeling approaches.