The cloud environment has various services at different levels of the architecture and the resources available at certain level of the architecture can be accessed through such service. The user data located at cloud data centers has to be provide public auditability where the access of data has to be restricted in many ways not only by encrypting the data at the clouds. There are many malicious threats been involved at any point of time to get access the cloud data in malformed manner or tries to spoil them by malware and so on. To overcome these issues of cloud security, we propose a service orient approximation technique which restricts the user access in many ways and uses multi attribute encryption standards for the cloud data to form an intrusion detection system to secure cloud environment from various threats. The user request being approximated based on the service requested and the approximation of the request is performed in many ways using number of factors like, the frequency of access, the user profile, the role in organization and many more. The proposed approach enforce an encryption standard based on number of attributes and the intrusion detection is performed using service approximation and encryption standards and user keys. Based on all these measures and variants the proposed approach maintains data auditability and provides efficient security to the cloud resources. The proposed method increases the data security, auditability and reduces the complexity in user authentication and increases the efficiency of intrusion detection. © 2015 Elsevier B.V., All rights reserved.