Enhancing Data Security and Privacy with a Modified Paillier Cryptosystem: A Lightweight Approach to Additive Homomorphic Encryption

The growing demand for secure and privacypreserving computation in modern applications, such as cloudbased data processing, secure voting systems, and healthcare analytics, has led to significant advances in homomorphic encryption techniques. This paper presents a modified implementation of the Paillier cryptosystem, a widely recognized additive homomorphic encryption scheme, tailored for enhanced security and performance. The proposed modification introduces efficient key generation using large primes, secure randomization techniques, and optimized modular arithmetic to ensure robust encryption without compromising computational feasibility. Additionally, the algorithm incorporates a padding mechanism based on cryptographic hashing to mitigate chosen-plaintext and frequency analysis attacks. We explore the theoretical foundation of the Paillier cryptosystem, its homomorphic properties, and practical considerations for its implementation, including scalability, latency, and memory efficiency. The proposed modifications ensure stronger inherent security while maintaining the simplicity and additive homomorphic capabilities of the original system. Experimental results validate the enhanced security and performance, making this approach suitable for lightweight and resourceconstrained environments.