Characterization of First and Second Order Statistics of Large Scale Fading Using Vehicular Sensors

We propose and demonstrate a framework for statistical modeling of propagation using spectrum sensors mounted on moving vehicles. To illustrate the concept, we deploy sensors on four vehicles to collect signal strength measurements together with corresponding locations on a suburban road. From the data we build a model for large scale fading, that is, the path loss and shadowing, as well as the shadowing correlation of a TV transmitter on two UHF channels. Depending on a specific band to which it is applied, such a model can be used for different purposes like coverage planning, to improve accuracy of a white space spectrum database, for selection of relay nodes in an ad- hoc or a mesh network, selection of spectrum sensors with significant diversity gain for cooperative sensing, etc. For instance, obtained numerical results imply that sufficient decorrelation of signal strength can be observed in the experimental area if the sensors are separated by more than 20m.