Accurate IoT device classification is essential for secure and efficient network management in heterogeneous environments. However, existing approaches struggle with overlapping traffic patterns and limited generalization across dynamic device behaviours. This paper proposes a Metadata Fusion with Attention-based BiLSTM (MFA-BiLSTM) framework that integrates time-domain and statistical features using an adaptive attention mechanism. The model captures both sequential dependencies and distributional characteristics of network traffic, enhancing feature representation and classification robustness. Experiments conducted on the CIC-IoT-Dataset2022 demonstrate improved performance with an accuracy of 99.40%, precision of 99.39%, recall of 99.36%, and F1-score of 99.38%. The results indicate that the proposed approach achieves reliable and scalable IoT device classification while maintaining interpretability.

Data fusion is generally defined as the application of methods that combines data from multiple sources and collect information in order to get conclusions. This paper analyzes the signalling time of different data fusion filter models available in the literature with the new community model. The signalling time is calculated based on the data transmission time and processing delay. These parameters reduce the signalling burden on master fusion filter and improves throughput. A comparison of signalling time of the existing data fusion models along with the new community model has also been presented in this paper. The results show that our community model incurs improvement with respect to the existing models in terms of signalling time.