This TransformerNano-communicationpaper presents a detailed transformerTransformer-based framework for analyzing traffic in electromagnetic in-body nano-communicationNano-communication networks with state of art models like BERT, RoBERTa, and ALBERT. These cutting-edge models are applied to improve the precision and reliability of traffic classification in the nano-communicationNano-communication field. Experimental results show that BERT achieves a training loss of 0.4292, validation loss of 0.4537, training accuracy of 79.25%, and validation accuracy of 78.00%. These findings therefore enhance the efficacy of our proposed framework for solving fundamental challenges in traffic classification that are integral to power biomedical applications such as real time healthMonitoring monitoringHealth monitoring, anomaly detectionAnomaly detection, and disease diagnosis. Future research topics include real time predictionPrediction of traffic, low power implantable sensors, and scalable solutions for large, dynamic biomedical databases.