A great deal of data is generated and collected by devices connected to healthcare systems on the Internet of Things (IoT). This data is distributed in real time, is unstructured in nature, and it is also a big problem to store and process in IoT applications. Since cloud computing is so common, many healthcare providers store these EHRs in cloud systems. Unauthorized access to these medical records is nevertheless still an issue. In order to avoid unauthorised access to this medical data in the public cloud networks, many access management models, along with cryptographic techniques have been used. This article proposes Enhanced Role Based Access Control (ERBAC) and Twofish algorithm to protect IoT health data in health systems from a public cloud storage perspective. The authors believe that the proposed system will significantly aid in the efficient storage of medical data in IoT applications and will provide secure storage of medical data in the cloud based on these role-based access policies. Additionally, a clustering strategy is introduced into the paper to reduce the waiting time for retrieving relevant medical data. The clustering of clinical data is the rational underlying approach for discovering secret examples from a wide variety of clinical data. Clinicians have rendered a professional decision on the disease probability using these cases. Clustering categories, the data set of separate collections based on comparability of information in the database is larger than other collections.Thus, this work proposes a clustering method that uses particle swarm optimization (PSO) with a genetic algorithm (GA) and proposes another clustering strategy called clustering calculation, based on the calculation of the advance of a swarm of molecules. It uses worldwide improvements in PSO calculation to fill the lack of the grouping strategy