Routing Optimization Based on Distributed Intelligent Network Softwarization for the Internet of Things

TL;DR

The paper tackles routing in IoT networks under energy and scalability constraints by introducing a distributed intelligent network softwarization framework that leverages SDN, NFV, and ML. It proposes Federated Deep Reinforcement Learning (FDRL) for routing, enabling multiple SDN controllers to train locally and share global policy updates via Federated Averaging, with a multi-objective reward that balances throughput, delay, loss, and hop count. The approach defines state, action, and reward structures for DRL, and demonstrates, through Mininet/Ryu-based simulations, that both distributed control and intelligent routing yield significant performance gains over centralized or conventional shortest-path routing, across multiple topologies. The work underscores the potential of privacy-preserving, scalable, distributed learning to enhance IoT routing and outlines future directions including NFV-enabled dynamic routing and security enhancements. $U(t)=w_{1}x_{t}-w_{2}d_{t}-w_{3}l_{t}-w_{4}h_{t}$, $A=igl\{a_{e} \,|\, e\in E\bigr\}$, and $G(V,E)$ are central to the formulation.

Abstract

The Internet of Things (IoT) establishes connectivity between billions of heterogeneous devices that provide a variety of essential everyday services. The IoT faces several challenges, including energy efficiency and scalability, that require consideration of enabling technologies such as network softwarization. This technology is an appropriate solution for IoT, leveraging Software Defined Networking (SDN) and Network Function Virtualization (NFV) as two main techniques, especially when combined with Machine Learning (ML). Although many efforts have been made to optimize routing in softwarized IoT, the existing solutions do not take advantage of distributed intelligence. In this paper, we propose to optimize routing in softwarized IoT networks using Federated Deep Reinforcement Learning (FDRL), where distributed network softwarization and intelligence (i.e., FDRL) join forces to improve routing in constrained IoT networks. Our proposal introduces the combination of two novelties (i.e., distributed controller design and intelligent routing) to meet the IoT requirements (mainly performance and energy efficiency). The simulation results confirm the effectiveness of our proposal compared to the conventional counterparts.

Figures (7)

• Figure 1: Distributed Intelligent Network Softwarization Architecture for IoT zormatimass23
• Figure 2: FDRL Routing for Softwarized IoT
• Figure 3: FDRL Training Phase
• Figure 4: Adopted Topology to Evaluate Distributed Control
• Figure 5: Variation of Delay with Time