The development of the edge devices and network is mainly depending of the trust of the user in the network. The vulnerable activities in the network affected the privacy of the data and minimize the trust of the users. The existing models failed to predict the privacy preserving ability of the edge devices due to limited prediction capability of traditional models. To overcome the existing drawbacks, the weight factor-based term frequency–inverse document frequency with sigmoid logistic regression (WFTFIDF-SLR) model was developed for the authentication of trust and privacy preservation of the edge devices. The data for the analysis were taken collected from the various contrast of the edge devices. The z-score normalization technique was utilized in the preprocessing phased for rescaling the ranges of the data. The weight factor-based term frequency–inverse document frequency (WFTFIDF) for the extraction of important details from the edge device data. The sigmoid logistic regression (SLR) algorithm for the authentication of trust and privacy preservation of the edge devices. The proposed WFTFIDF-SLR model has shown upgraded results with accuracy of 96.13%, f1-score of 95.19%, recall of 95.89%, and precision of 94.51% compared to the existing privacy protection-based federated deep learning (PP-FDL).