Design and Implementation of a Lightweight Bloom Filter Accelerator for IoT Applications

The Internet of Things (IoT) has evolved into a rapidly expanding web of devices surrounding our lives. This interconnected ecosystem facilitates a new level of collaboration and the collection of precious data to be distributed and analyzed. However, as the number of connected devices grows, the volume and variety of generated data increase significantly, necessitating efficient and scalable solutions.Membership testing is one of the challenging the algorithms used in most of the network applications. To make it fast and lightweight, probabilistic data structures like Bloom Filters are employed. Although variants of this algorithm are proposed in the literature, there are very few addressing low power, low resource implementations with recently developed hashing techniques. Therefore, in this work, we present a practical Bloom filter accelerator design with Murmur3 hash, implemented on a Nexys A7 FPGA board. Then, we test and analyze its performance to identify limitations and explore possible improvements.