The in-network processing consists of data aggregation operations e.g. Summation, averaging, min-max value, variance etc. Data aggregation reduces the number of message transmitted to the Wireless Sensor Network(WSN) by calculating the sub aggregation results at the intermediate nodes. Furthermore the sensor nodes are deployed in open and unsafe environments, so the security of sensed and aggregated data is crucial. This situation necessitates the investigation of the Secure Data Aggregation (SDA) protocol. Many protocols have been proposed concerning finding answers for SUM queries in WSNs but most of them either offer only approximate answers for SUM queries or do not live up to all the security requirements. The focus of the research in this paper is to investigate a SDA protocol which satisfies all the security requirements viz. Confidentiality, Privacy, Authentication, Integrity and Freshness of the query result and also returns the exact answers for SUM queries (as well as their derivatives, e.g., COUNT, AVG, etc). The protocol applies homomorphic encryption that promises the privacy as well as confidentiality of data sent by sensor nodes with innetwork aggregation. The protocol achieves integrity of sensed data by means of secret sharing scheme, message authentication code (MAC) and data diffusion methods. Proposed scheme satisfies all the essential security requirements for secure in-network aggregation. This scheme does not impose extra overhead in communication. Therefore, the scheme is best desirable for resource-restrain WSNs.