<p>The popularity of IoT devices has been on the rise, with their integration into various applications, including smart home systems. IoT devices communicate over the network, which an invader can snoop on. The user does not need to rely solely on the unencrypted traffic, as user activities can be derived from encrypted content. Traffic privacy should be considered a primary concern, especially in smart homes. Traffic shaping techniques may be applied to obscure the data flow, preventing meaningful inferences through traffic analysis. Existing traffic shaping approaches, however, involve multiple configurable parameters that are challenging to fine-tune for achieving a balance between bandwidth consumption and response times. Therefore, in this study, we analyse current traffic shaping algorithms regarding their computational demands, bandwidth usage, latency effects, and privacy safeguards, using network traffic data collected from a simulated smart home environment. A novel traffic shaping approach, Dynamic Traffic Padding (DTP), is introduced to optimize the trade-off between bandwidth use and latency based on device types and the desired level of privacy. The proposed approach is based on prior device usage activity and takes reasonable measures to reduce bandwidth usage from the general traffic pattern. The effectiveness of the introduced traffic holding method in a coherent manner sustains data from the virtual smart home based on additional suggestions from other variables.</p>