A Measurement-Based Model for Dynamic Spectrum Access in WLAN Channels

In this paper we consider dynamically sharing the spectrum in the time-domain by exploiting whitespace between the bursty transmissions of a set of users, represented by an 802.11b based wireless LAN (WLAN). Realizing that exploiting the under-utilization of the channel requires a good model of the these users' medium access, we propose a continuous-time semi-Markov model that captures the WLAN's behavior yet remains tractable enough to be used for deriving optimal control strategies within a decision-theoretic framework. Our model is based on actual measurements in the 2.4 GHz ISM band using a vector signal analyzer to collect complex baseband data. We explore two different sensing strategies to identify spectrum opportunities depending on whether the primary user's transmission scheme is known. The collected data is used to statistically characterize the idle and busy periods of the channel. Furthermore, we show that a continuous-time semi-Markov model is able to capture the data with good accuracy. The Kolmogorov-Smirnov test is used to validate the model and to assess the model's goodness-of-fit quantitatively. A conclusion summarizes the main results of the paper