It is becoming very convenient for Internet users across the globe to have at their fingertips major online services, such as banking, equity share marketing, medical information, shopping, and much more. It is important to consider user privacy and security while performing these online activities, especially if the application is highly sensitive, such as online banking. While accessing these online applications, most users are worried about Internet vulnerabilities and thus try to prevent assault by intruders/hackers and ensure that their transactions are secure with Internet aids such as a firewall, anti-phishing software, anti-virus software, and data encryption/decryption capabilities, to name a few. In spite of taking these precautionary measures, recent research has revealed users' activity being compromised by intruders who use a side-channel attack by studying the packet's timing information, packet size, and packet direction, which is facilitated by an application's convenience features such as auto-suggestion and auto-completion. These features result in exhibiting unique traffic patterns for each character entered and thus enable a side-channel attack. In this thesis, a method to counter these attacks is provided by employing a packet-padding mechanism, thus ensuring that each individual packet is similar to enough other packets in size and orientation. Compared to other existing solutions, this method provides an intelligent solution with minimum padding cost and lower network bandwidth overhead.