Implementing an Efficient Architecture for Latency Optimisation in Smart Farming

In various industries, including Agriculture, the application of the Fifth Generation (5G) of wireless technology has led to significant advancement. One of the most intriguing aspects of 5G technology is the potential to reduce latency for Internet of Things (IoT) applications, especially for latency-sensitive smart farming applications such as fire detection. The data generated by IoT devices such as sensors, cameras, and actuators in intelligent farming applications are growing exponentially. Traditional methods of processing and storing IoT data usually include cloud data centres, which are often far from data sources, resulting in multiple network hops that increase latency. To this end, the existing network infrastructures struggle to cope with increasing traffic and meet the stringent low-latency requirements of various real-time environmental monitoring IoT applications. To solve this problem, Edge Computing (EC) emerged as a solution. Edge technology allows the deployment of 5G Core (5GC) network functions close to the data-generating IoT sensors. This method allows data processing to be performed near the sensor, thereby reducing latency. As a result, this paper proposed an efficient architecture to minimise latency in 5G networks by moving the User Plane Function (UPF) node to the edge of the network closer to users using the Control and User Plane Separation (CUPS) strategy. Furthermore, this paper further proposed Software-Defined Networking (SDN) based backhaul. This backhaul was configured to use the Open Network Operating System (ONOS) controller, which has been customised with a distributed core to improve throughput, latency, and scalability. Using SDN in the 5G backhaul network allows operators to create dynamic, scalable, and efficient networks capable of serving a diverse variety of services and applications with varying performance needs. The results of the experiment conducted on the Third Generation Partnership Project (3GPP) compliant 5G testbed demonstrated that the proposed architecture reduced the average Round Time Trip (RTT) by 60.7% while improving the throughput by approximately 40.48%. This significant reduction in latency and improvement in throughput paves the way for the implementation of real-world 5G applications, particularly in latency-sensitive sectors like smart agriculture. Agricultural operations will benefit from faster data processing and communication, enabling real-time monitoring, and precision irrigation, u