Hardware based cyber system using high performance crypto hash bloom filter for network security and privacy preserving applications

Abstract

Bloom filter is an excellent data structure mechanism for any system to perform membership querying portably. Conventional Bloom filters use universal hash functions with notable false positive ratio due to hash collisions and yields performance degradation. Several cryptographic hash functions were surveyed for possible incorporation in Bloom filter deployments. Suitable and efficient hash functions like Tiger hash and Whirlpool hash which has more enhanced security and suitability features are chosen. Tiger hash is a family of Merkley Damgard designed for high speed with less collision yielding security and speed in applications. Whereas, the whirlpool hash function is a Standardized hash constructed based on Advance Encryption Standard (AES) and its status is updated by a four-round function by using unique encryption process. Novel high-performance bloom filters using the selected cryptographic hashing schemes can be developed for secure indexing of member elements. Tiger hash is selected in this book chapter for discussion. Bloom filter is modeledto be useful in various network security and querying applications like Intrusion detection systems, pattern matching, packet classification, detection of the flooded attack on the internet, distributed caching of web servers, etc., This new modeling requires suitable modifications in standard Bloom filter and also in the hash functions. In this chapter of the book, we discuss specifically the case of Crypto Hash Bloom filter incorporation in Network Intrusion Detection System (NIDS) in security applications and Privacy-preserving record linkage in medical research applications and show the improved and comparable performance of the proposed system. The hardware implementation of the variants in various FPGA devices also discussed. This work provides better results in applications than the previously existing systems and paves the way for huge future research scope for the research community.