AoI in Context-Aware Hybrid Radio-Optical IoT Networks

TL;DR

This work tackles timely data delivery in IoT by leveraging a hybrid RF-OC network and a multi-objective optimization that jointly maximizes throughput, minimizes energy, and reduces technology switching. It introduces decision variables for transmission, scheduling, and technology choice, and applies Big-M linearization to handle non-linear constraints within a convex MINLP framework. Simulation results demonstrate that integrating Optical Communication with RF significantly lowers Mean AoI and Peak AoI compared with RF alone, indicating fresher data collection under realistic energy and delay considerations. The approach provides a principled method to adaptively select the best communication technology per device, improving data freshness in dynamic IoT environments with acceptable computational overhead. Practical impact includes more responsive IoT systems for healthcare, smart environments, and other time-sensitive applications where timely information is critical.

Abstract

With the surge in IoT devices ranging from wearables to smart homes, prompt transmission is crucial. The Age of Information (AoI) emerges as a critical metric in this context, representing the freshness of the information transmitted across the network. This paper studies hybrid IoT networks that employ Optical Communication (OC) as a reinforcement medium to Radio Frequency (RF). We formulate a non-linear convex optimization that adopts a multi-objective optimization strategy to dynamically schedule the communication between devices and select their corresponding communication technology, aiming to balance the maximization of network throughput with the minimization of energy usage and the frequency of switching between technologies. To mitigate the impact of dominant sub-objectives and their scale disparity, the designed approach employs a regularization method that approximates adequate sub-objective scaling weights. Simulation results show that the OC supplementary integration alongside RF enhances the network's overall performances and significantly reduces the Mean AoI and Peak AoI, allowing the collection of the freshest possible data using the best available communication technology.

Figures (6)

• Figure 1: System Model. This example shows the network using RF-OC to exchange 3 different types of applications' data.
• Figure 2: Avg. consumed energy (%) vs Avg. Delay (%), For RF and OC systems. The figure shows the feasible regularization $\alpha_1$ for each standalone system. The star represents the values of $\alpha_1$ with a delay of zero for the assigned communication.
• Figure 3: AoI vs. Number of APs (a) and AoI vs. Number of IoT nodes (b) for RF and Hybrid RF-OC systems.
• Figure 4: AoI for RF (a) and AoI for RF-OC (b) vs. time.
• Figure 5: Avg. Trans. Rate (a) and Avg. Consumed Energy (b) vs. Number of IoT devices for RF and RF-OC with $N_{APs}=3$.