Packet filtering techniques play an important role in many of network devices such as firewalls, IPSec Gateways. Firewall plays an important role in safeguarding any system from any external attacks to the system. It can be used to safeguard hosts as well as networks. This research focuses on studying the performance impact and the sensitivity of the Linux firewall (IP tables) also improve by using this research. And these are improving to become fast. A firewall designed in Linux, the user can edit the source code and change it depending on the security requirements for the LAN. At any time one can configure the firewall to encrypt, to decrypt, accept, deny, or proxy all packets that are being sent between any two systems depending on the rules. On the basis of this, the user can be blocked or given access to a network using a good tree algorithm. There are two approaches for the filtering, first by using the early rejection of unwanted flows without impacting other flows significantly. Second, we present a new packet filtering optimization technique that uses adaptive statistical search trees to utilize important traffic characteristics and minimize the average packet matching time. The proposed techniques timely adapt to changes in the traffic conditions by performing simple calculations for optimizing the search data structure. The proposed techniques can significantly minimize the packet filtering time with reasonable memory space requirements.