International Journal of Engineering Practical Research (IJEPR) Volume 2 Issue 1, February 2013 www.seipub.org/ijepr 1 Ku‐band Channel Signal Generator Based on a Statistical Channel Model Sánchez‐Salas, D.A. 1 , Cuevas‐Ruíz, J.L. 2 1 Department of Mechatronics Engineering, Tecnológico de Monterrey, Campus Estado de México, México 1 Department of Technology Information and Electronics, Tecnológico de Monterrey, Campus Santa Fe, México Emails: A00465173@itesm.mx; jose.cuevas@itesm.mx Abstract Some wireless channel models based on probability density functions have been proposed, including models based on Markov chains. However, they are limited to a certain number of perturbations. This paper proposes a methodology based in a Ku‐band signal behavior which is classified in three cases: ascendant, descendent and constant. Samples are classified in these cases so their conditional probabilities are analyzed in order to find a fit to a probability density function and to extract its statistical parameters which are the model itself. From this information, a new signal was generated and their second order statistics are compared with the ones from the original signal to validate the created model. The analyzed signal was extracted from a measurement campaign done in Mexico. Keywords Ku Band Channel Model; Statistical Channel Model; Signal Generator; Second Order Statistics Introduction A deep knowledge is required in the design and optimization of wireless communication systems. Models based on probability density functions (pdf) are not enough to characterize a dynamic channel because they only describe one type of phenomenon so other alternatives are analyzed. Many publications describe models created from measured samples where the signal is analyzed as a random variable and only one perturbation is characterized or it can only be applied to certain frequency or system. Other models are based in Markov chains which implies the combination of many phenomena however they are limited to a pre‐established number of states or events during the transmission [(Barts, 1988), (Fontan, 2001), (Abouraddy, 2000), (Kattenbach, 2002), (Eberlein, 2007)]. Another alternative for channel modeling is to analyze the variations in the level of the signal due to all phenomena (like rain, scintillation, etc.) which produces its fading. These variations are denominated as the dynamic of the wireless signal. The classification according to this dynamic is suggested by U.C. Fiebig in (Fiebig, U.-C, 1999) where samples which contain the attenuation of a signal are classified in three cases: ascending, descending and constant. An example of this classification is shown in FIG. 1. A wireless channel model for a set of samples from a Ku band signal is proposed in this paper. The model is based on works of (Fiebig, 1999) where a classification of the samples is done according to its dynamic, next a goodness of fit test is done to know which pdf fit best. The objective is to extract the statistical parameters of the fitted pdf for every value of level of the signal and case. These data are the model itself and a new time series can be generated from them. The validation of that process is done by comparing the second order statistics (average fading duration –AFD ‐ and level crossing rate ‐ LCR) of the original signal and the new generated signal. The document is organized as follows: the measurement campaign organized to extract the samples of the signal is described in section II; in section III the methodology is explained and it will be applied to data in order to obtain a new signal; results are explained in section IV; conclusions are given in section V. FIG. 1 CLASSIFICATION OF SAMPLES ACCORDING TO ITS DYNAMIC