Reducing routing table size using ternary-CAM

Ternary Content Addressable Memory (TCAM) has increasingly been used in high speed routers to perform routing lookup function. They allow simultaneous comparison of the key with every index at the same time so that the longest matched prefix could be selected within one memory access, much faster than software based search algorithms. We propose two techniques to compact routing table stored in TCAM. The techniques allow a smaller TCAM to be used to reduce cost, power consumption and thermal dissipation. They could also help routers to scale to larger routing table. Our simulation shows up to 48% size reduction. Fast incremental update algorithm is presented to maintain space saving without recomputing the compacted table after each routing update.