Access-efficient Balanced Bloom Filters

Yossi Kanizo, David Hay, Isaac Keslassy
2012 2012 IEEE International Conference on Communications (ICC)  
The implementation of Bloom Filters in network devices ideally has low memory access-rate and low false positive rate. However, since Bloom Filters hash elements to arbitrary memory blocks, they need high memory access rates. On the other hand, Blocked Bloom Filters first hash elements to a single memory block, where they maintain a local Bloom Filter. Therefore, they access only one memory block per element, resulting in better memory-access efficiency. Unfortunately, they have poor
more » ... ave poor performance, with poor load-balancing and high false positive rates. In this paper, we propose to implement load-balancing schemes for the choice of the memory block, along with an optional overflow list, resulting in improved false positive rates while keeping a high memory-access efficiency. To study this problem, we define, analyze and solve a fundamental access-constrained balancing problem, where incoming elements need to be optimally balanced across resources while satisfying average and instantaneous constraints on the number of memory accesses associated with checking the current load of the resources. We then build on this problem to suggest a new load-balanced Blocked Bloom Filter scheme. Finally, we show that this scheme can reduce the false positive rate by up to two orders of magnitude, with the cost of 1.2 memory accesses per element and an overflow list size of 0.5% of the elements.
doi:10.1109/icc.2012.6363636 dblp:conf/icc/KanizoHK12 fatcat:gpcod342v5dvlhkihg5yp2tn7m