This paper proposes a new efficient packet classification algorithm based on area-based quad-trie (AQT) algorithm. The AQT is an efficient algorithm in memory requirement since rule replication is not occurred. But it does not provide high-speed search because of linear search for multiple rules stored in a node of the quad-trie. Boundary cutting (BC) algorithm provides high-speed search, but it has a rule replication problem; memory requirement is very high. This paper is motivated to improve the search performance of the AQT by applying the boundary cutting for nodes that has more rules than a pre-defined number. Simulation results show that the proposed algorithm considerably improves the search performance of the AQT and has lower memory requirement than the BC.