The need for latency-sensitive applications in fog-based cloud computing has significantly increased as a result of the current Internet of Things (IoT) paradigm. These applications, however, are unable to satisfy stringent quality of service (QoS) standards. To guarantee QoS when processing huge data, the extensive implementation of IoT necessitates more efficient use of network infrastructure. Deploying cloud-centric IoT applications often entails several modules operating on cloud servers and terminal devices. It is challenging to implement latency-sensitive applications in a heterogeneous fog computing environment as fog devices with varying processing capacities must handle the data produced by the end device. Furthermore, the implementation of such apps becomes more complex when the fog and the terminal device have uneven connection delays. In this paper, we provide an approach that may efficiently allocate application modules to network nodes while taking processing power, connection latency, and sensing data volume into account. To verify the algorithm's efficacy, we ran simulations in iFogSim in comparison to a standard cloud computing deployment. The efficiency of the suggested method in terms of end-to-end latency and network utilisation is confirmed by the simulation results. In this case, the number of sensors has no effect on latency or execution time.

Internet of Things, fog-cloud computing architecture, resource-aware placement, latency-sensitive module placement, and application