The security of Wireless Sensor Network (WSN) in its applications face great challenges while satisfying energy constraints in green computing. WSN have different kinds of constraints such as lack of infrastructure, limited memory, low computation capability and reduced battery life, by those, security implied as supplemental demanding job. Since, there is much difference in reading values of sensor node and failure in nodes; it is difficult for accurate data extraction and aggregation in WSN. The principal purpose of data aggregation (DA) is to collect data in an energy efficient manner by which the network existence will be increased. This model investigates security linked issues, challenges and requirements in WSN. And also, a trust based framework for predicting the effectiveness of secured DA in WSN is presented. In this technique, a Trust rating function is developed to find the trust rating value of each sensor nodes and then form data aggregation tree (DAT) based on modified Cuckoo search algorithm (CSA). The sensor nodes with higher trust rating are selected as aggregators. Each sensor node encrypts data using watermarking techniques. Encrypted data is aggregated and transmitted to sink by aggregator. The simulation outcome depict that the intended model have superior than the existing one.