<p>The number of devices connected to the Internet is continually rising due to the expansion of the Internet of Things (IoT). Real-time IoT applications face significant challenges related to delays in service delivery from cloud computing resources, primarily due to limited bandwidth. Fog computing, which is more geographically dispersed and closer to end-users, enhances system efficiency by reducing data transmission distances, making it suitable for various delay-sensitive IoT applications. However, the limited resources of fog devices restrict their ability to deploy multiple applications effectively due to inadequate resource estimation and discovery mechanisms for diverse IoT needs. The heterogeneous computational capabilities of fog nodes and varying Quality of Service (QoS) requirements complicate application placement in fog environments. To address these challenges, an effective resource allocation strategy is crucial for meeting QoS requirements and improving overall system performance. Finding the optimal placement strategy for IoT applications with multiple QoS parameters is complex and classified as NP-complete. This paper focuses on using a population-based meta-heuristic algorithm inspired by Harris Hawks optimization to monitor QoS needs and available fog node capabilities, facilitating effective service placement. The proposed approach optimizes throughput, energy consumption, and costs while satisfying the QoS requirements of each IoT service. Simulation results indicate that this solution enhances resource utilization and service acceptance rates by 4.5% and 3.8%, respectively, while reducing service delay and energy consumption by 2.95% and 1.62% compared to other existing methods.</p>