In wireless communication environments such as indoors and in a relatively narrow region for which applications have recently started, the information transmission speed is sufficiently faster than the fading variation. In such cases, the fading variation can be viewed as a colored probability process with autocorrelation. In the present paper, as a method of forming a fading environment model, a successive method for fading coefficients by the Jakes model based on the probability configuration is presented. Further, a maximum likelihood estimation method using a conditional probability density is proposed for deriving the fading coefficients. As an application example, the CCK method, which communicates high‐speed data of a wireless LAN under the IEEE 802.11b standard, is studied. The decoding characteristics in a fading environment are evaluated by simulation, and it is found that an SNR improvement of about 1 dB is possible for an error rate of 10−3 level in comparison with the conventional polynomial interpolation method. © 2005 Wiley Periodicals, Inc. Electron Comm Jpn Pt 3, 88(5): 1–11, 2005; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ecjc.20152