Since vehicular channels are non-stationary wireless channels, should they be modeled as linear or non-linear systems? If non-linear what is the source of non-linearity?
In a wireless link, the signal propagates from the Tx (transmitter) to the Rx (receiver) via several propagation paths. The contributions of those paths add up at the Rx.
As a result we experience fading, variation of the received signal power with time, and signal echoes with different delays. The contributions of the various propagation paths, i.e., amplitudes and phases, and their respective delays
define the impulse response. The expected (i.e., averaged) power at different delays is described by the average power delay profile (APDP), from which the so-called root mean square (rms) delay spread is evaluated as the second central moment. For vehicular channels, it is customary to distinguish between V2V and V2I channels. These channels not only differ from each other, but also deviate significantly from those in cellular communication. All these properties influence the input–output relationship of the channel, i.e., how signals propagate from the transmitters to the intended, and other nonintended, receivers. Since these relationships vary between different scenarios, it is highly unlikely that a wireless system optimized for one specific scenario will also work well in other, completely different, scenarios. Due to the high speeds involved, V2V and V2I channels show strong time variance (the channel state changes) and nonstationarity
(the channel statistics change). These effects are more pronounced for cars approaching each other or approaching intersections, while they are less severe for vehicles driving in convoys, or V2I communications.
The system is termed linear if the response to weighted linear combination
of two (or more) signals is the same weighted combination of responses to the two (or more) signals, for all possible choices of x1[.], x2[.] a1 and a2, i.e., if y3[n] = a1 y1[n] + a2 y2[n] for all n ,where yi[.] denotes the response of the system to input xi[.] for i = 1, 2, 3. So you can find the answer in the attached paper.