A Mininet-Based SD-WLAN Emulator for Hybrid LiFi and WiFi Networks with Performance Analysis

Combining the high-speed transmission of light fidelity (LiFi) and the ubiquitous coverage of wireless fidelity (WiFi), hybrid LiFi and WiFi networks (HLWNets) have been recognized as a promising technology for next-generation wireless local area network (WLAN). For such a hybrid network, load balancing (LB) is essential and crucial for exploiting the full potential of network performance. So far, an extensive number of relevant studies have been carried out, exploring various approaches from game theory to machine learning. However, existing research is mainly based on simulations of network capacity, neglecting the impact of realistic network protocols and operations. In this paper, we develop the first ever network emulator for HLWNets, which is based on Mininet while incorporating software-defined networking (SDN) to realize the functionality of LB. The emulator consists of four key components: network deployment, LB decision making, LB implementation, and real-time performance measurement. With the constructed emulator, the performance of several representative LB methods is evaluated, including signal strength strategy (SSS), distributed optimization (DO), and game theory (GT). It is found that with respect to DO and GT, there exists a gap of about 10 % between the simulated network capacity and the emulator measurement results. In terms of SSS, this gap is less pronounced but still exceeds 4 %. These results prove the importance of the network emulator when studying LB for HLWNets and other similar heterogeneous networks.