Realistic and Efficient Channel Modeling for Vehicular Networks

摘要: To date, VANET research efforts have relied heavily on simulations, due to prohibitive costs of deploying real world testbeds. Existing channel models implemented in discrete-event simulators are by and large simple stochastic radio models, based the statistical properties chosen environment. It was shown that such unable provide satisfactory accuracy for typical scenarios. We performed extensive measurements different environments (open space, highway, suburban, urban, parking lot) characterize detail impact vehicles communication terms received power, packet delivery rate, effective range. Since found be significant, we developed a model accounts as three-dimensional obstacles takes into account their line sight obstruction, signal reception message reachability. The is empirically derived vehicle dimensions, accurate positioning, realistic mobility patterns. We validate against measurements. enable modeling urban suburban environments, incorporates static objects well. requires minimum geographic information: location dimensions modeled (vehicles, buildings, foliage). show it successfully captures both small-scale large-scale propagation effects (highway, open space).