The Internet of Things (IoT) expansion has increased in sectors like health care, industrial automation, and smart cities. But with this growth in devices comes growing worries about data security and privacy. However, traditional cryptographic methods tend to be inadequate for IoT environments that demand ciphertexts to be decrypted for processing, uncovering sensitive information in the process, thus expanding vulnerability in IoT networks. In this work, we investigate how to tackle these challenges using homomorphic encryption, particularly the Paillier cryptosystem, which enables computations on encrypted data without decryption. Our approach is to protect data in IoT from the point of collection until processing and storage. In this research, we extensively study the Paillier cryptosystem: its crucial generation, encryption, decryption mechanism, and unique homomorphic properties, which support secure operation on encrypted data. The encryption and decryption times, data transmission latency, and CPU and memory resource utilization were evaluated extensively in a simulated IoT environment. Comparisons with other cryptographic approaches show that Paillier encryption balances security and computational efficiency and is well-suited for resource-constrained IoT devices. Our results indicate that the Paillier cryptosystem improves data confidentiality and security in IoT networks and allows secure real-time data processing without incurring a performance penalty on the device. At the end of this study, we conclude that although Paillier encryption is a promising solution, additional optimization and the use of more powerful homomorphic schemes can allow it to be used in more comprehensive IoT security frameworks.