Recommendations for shadow fading estimation from received composite signal samples

Evident lack of verified and comprehensive theoretical models for shadow fading, which enable analytical estimation of its statistical distribution and statistical parameters, has promoted the empirical alternatives for this task in many aspects of a wireless system lifecycle, such as e.g. coverage planning, where the imperative is to determine the optimal number of base stations, which balances the system performance and implementation efficiency. However, in addition to shadow fading, the composite received signal is simultaneously affected by deterministic path loss and multipath fading, too, so the prerequisite for accurate estimation of shadow fading from the composite received signal samples is efficient elimination of its fast temporal variations, fast spatial variations, and finally, path loss. Unfortunately, the available methods for this are often not appropriate for measurements related to shadow fading estimation. So, for example, even the commonly used drive/walk test provides neither elimination of time variations, nor accurate path loss estimation. So, in this paper, after identifying drawbacks of the existing techniques for shadow fading estimation, the appropriate procedure is proposed with concrete and unambiguous guidelines for elimination of undesirable components of the composite signal through a five-step algorithm, which is shown to provide significant advantage with respect to common drive/walk testing, in terms of shadow fading values and its statistical parameters estimation.