Primarily job of a node in Wireless Sensor Networks (WSNs) is to collect the data, process it across various applications and send it to the appropriate location. Ensuring the integrity and reliability of data is a significant challenge due to complex and unpredictable operational environments of WSN in which they operate. This paper emphasizes the crucial role of anomaly detection in sensor data, acting as a fundamental element for bolstering data integrity and fortifying the resilience of WSNs. Our methodology revolves around the utilization of Principal Component Analysis (PCA), a potent technique for reducing data dimensionality while retaining essential features, resulting in a refined dataset. This refined data forms the basis for training a Decision Tree model, known for its ability to discern between normal and anomalous data patterns. As a result, our approach provides a robust and interpretable framework for anomaly detection within WSNs. The potential of our proposed approach spans a broad spectrum of domains, encompassing applications in environmental monitoring, healthcare, and industrial automation. This approach holds the promise of enhancing the performance and effectiveness of WSNs across a diverse range of applications.