Privacy-Preserving Phrase Search Algorithm Using Bloom Filters and Hybrid AES_GCM for Secured IoT Data in Cloud Environments

Abstract

The growth of IoT devices creates huge amounts of sensitive data that are usually sent to cloud platforms for storage and processing. But making it possible to do a secure and quick phrase search on encrypted IoT data is still hard because of privacy concerns and limited computing power. This paper suggests a new phrase search algorithm that protects privacy by combining Bloom Filters and a hybrid AES_GCM encryption scheme. Space-efficient, probabilistic search indexes are made with Bloom Filters that can be used to match multiple keywords and phrases with minimal additional storage space and false-positive rates. The AESGCM hybrid module secures the privacy and integrity of data by authenticated encryption. This ensures that the plaintext is less likely to be leaked or ciphertext will be altered when storing or transmitting the data in the cloud. Strict performance test and evaluation indicate that the suggested algorithm improves search accuracy, query response time, and computation and efficiency significantly than the conventional encrypted search techniques. Also, the scheme is very hard to break with inference and frequency-based attacks, which makes IoT-generated datasets even more private from end to end. The framework can handle dynamic updates and works well with different types of IoT data streams and queries of different levels of difficulty.  The work addresses essential security gaps in data searchability outsourcing and provides an easy and convenient method of obtaining data as well as guarding privacy in IoT-cloud environment with limited resources. It also preconditions the further safe IoT data handling designs.