Study of a hardware Hash list compression method and its performance analysis
Yong Zhang · Journal of Chongqing University of Posts and Telecommunications · 2004
In high speed and large volume flow of packet classification, hardware Hash table has the advantage of lower cost and better performance. In these applications, the flow ID should be stored in the hardware hash table. While in some applications, the flow ID can be very long, e.g. 300 bits, and it requires a much larger memory to hold these flow ID. At the same time, the Hash table look-up performance is degraded remarkably. In this paper, we proposed a hardware Hash list compression method. It can effectively reduce the flow ID length and Hash table memory size needed and improves the table look-up performance. The implementation of this method is very simple and it has much merit in practical applications.