Modeling wireless infrared communications for network simulation

Although optical wireless communications offers many advantages over radio communications for local area networks, there are few statistical models or simulation packages dealing with infrared wireless communications. In order to study the feasibility of infrared wireless local area networks, computationally efficient models are needed that accurately represent the infrared wireless channel, and these must be integrated with network simulation tools.; We first investigate the behavior of infrared channels inside empty rooms of various sizes. We collect thousands of impulse responses using a detailed propagation modeling tool and analyze the channel gain and delay spread to look for general trends. The delay spread is used to characterize the temporal dispersion of an impulse response. We discover that line-of-sight and diffuse channels behave differently when the receiver and transmitter are separated by a small distance (less than 2.0 m). For larger distances, line-of-sight channels have an average of 6 dB higher channel gain and 6 ns lower delay spread compared to diffuse channels. We find statistical distributions to fit channel gains and delay spreads for line-of-sight and diffuse channels. We show that all of our models will be accurate for data rates up to 100 Mb/s. The empty room tests were repeated for varying reflectivities of the room. Additional tests were carried out for furnished rooms. The previously developed models can be adjusted for reflectivity and furniture effects.; We compare our models with previously measured channel data and find them to be consistent for all channel types. These models are next integrated with GloMoSim, a wireless ad-hoc simulation tool designed for large scale simulation. In addition, the detection of walls and ambient noise sources is created in GloMoSim. These features are key to evaluating an infrared wireless communication system and do not exist in any current wireless simulation packages. Next, we evaluate the performance of sample wireless infrared networks using GloMoSim. In particular we look at achievable throughput for our different models of infrared propagation. This complete infrared wireless network simulation tool can be used to create and compare a multitude of possible networks.