One Time Trapdoor based Searchable Encryption

Data outsourcing has emerged as an effective solution to handle huge volume of data, with low costs, robustness and support for efficient analytics. But it also raises concern on data privacy. To overcome the privacy issue, data is stored in an encrypted format. However encryption makes search over the protected data convoluted. The need of the hour is to provide an efficient method for data storage and search over protected data at remote provider without compromising with security. In keyword based search, each data item is associated with a set of keywords over which the search is performed. For a search, user supplies keywords to the owner to get a corresponding trapdoor to be searched over the provider. But the trapdoor is vulnerable to online/offline keyword guessing attack. This paper proposes one time trapdoor scheme, where new trapdoors are generated every time for searching keywords, eliminates the risk of online and offline keyword guessing attack. The data encryption and trapdoor generation is based on the non-polynomial hard problem of bilinear pairing and discrete logarithm. We have considered health care data stored in hierarchical structure as a case study but the scheme holds good for other pattern of storage as well.