Replication of the Bursty Behavior of Indoor WLAN Channels

In this paper we present the design, implementation and assessment (by means of extensive simulation campaigns) of two wireless error models within the ns-3 framework, whose main goal is to emulate the bursty behavior of indoor real propagation environments. The first one, called Bursty Error model based on an Auto Regressive filter (BEAR), aims to mimic, including a memory-aware contribution, the received Signal to Noise Ratio, which is afterwards used to establish the presence of errors within the frame. The second one is an extension of the widespread Gilbert-Elliot channel model, based on a Hidden Markov Process. Both of them are proved to accurately replicate the presence of long error frame bursts, as compared to the legacy approaches which are integrated in the simulator, by studying a wide range of performance figures. The paper also discusses the drawbacks exhibited by the legacy error models supported by the simulator, according to the behavior observed over real indoor wireless channels, since they lead to unrealistic results.