Deterministic Models of the Physical Layer through Signal Simulation

Abstract

Current wireless network simulators provide very detailed and deterministic models of the network protocol layers, whereas rather simple and stochastic models, based on the signal-to-noise ratio, are used for the simulation of the physical layer. Although this approach can be sufficient to study the behavior of different upper layer protocol variations, it prevents an easy alteration of the physical layer because a stochastic abstraction of the physical layer must be provided in advance. In particular, the simulation of distributed systems with physical layers that are designed to have several senders transmitting signals at the same time intentionally, is hardly possible with current approaches. A further problem of stochastic physical layer simulations is the fact that the radio channel's influence must also be carried out stochastically, which limits the advantage of accurate rayoptical channel models. We present a novel approach for the accurate simulation of the physical layer by utilizing existing software-defined radio implementations to create signals, to calculate interference and to decode signals. This technique enables us to simulate wireless networks holistically and, furthermore, we can fully exploit the possibilities of available ray-optical channel models.

Performance Assessment of IEEE 802.11p with an Open Source SDR-Based Prototype

“…There is, however, also the idea to model the physical layer on signal level [11] and, thus, to combine physical layer and network layer simulators. In fact, our initial SDR implementation has already been integrated into Veins by an independent group [19]. Finally, trace-based input modeling can be used.…”

“…1 We believe that our SDR-based IEEE 802.11p transceiver will serve as a valuable tool to study many aspects of VANETs. In fact, it has already been used for privacyrelated research, where it provided access to information that is not available with normal WLAN cards [34], and large-scale vehicular network simulations, where it provided a realistic physical layer simulation model [19]. We think, however, that its most interesting applications in the signal processing context have yet to be explored.…”