Radio channel fingerprint model based on entropy weight method

Different radio channels have various characteristics, which is analogous to human “fingerprint”, and the difference between the characteristics is defined as radio channel fingerprint. In this paper, we construct a quantitative radio channel fingerprint model based on the entropy weight method. With this model, five indicators, including time delay spread, coherence bandwidth, Doppler spread, fading threshold, and level variance, are selected to characterize the channel fingerprint and generate the evaluation matrix. Then we apply entropy weight method to find the reasonable weights for the five indicators, and thus achieve quantitative radio channel fingerprint model. Finally, the proposed fingerprint model is used to correctly realize scene recognition with the practical measured data provided by Huawei Corporation. This success validates the effectiveness and correctness of the proposed model.