Design, modeling, and optimization of hybrid LiFi-WiFi networks for high-performance indoor wireless communication

Abstract

The exponential growth of bandwidth-intensive applications such as ultra-high definition (UHD) video streaming, cloud computing, augmented and virtual reality (AR/VR), and the Internet of Things (IOT) has pushed existing wireless communication technologies to their limits. Conventional RF-based networks, such as Wi-Fi and LTE, are increasingly strained by spectrum scarcity and escalating user demands. This has spurred interest in alternative or complementary communication paradigms capable of delivering higher data rates, lower latency, improved security, and better energy efficiency. Light Fidelity (Li-Fi), operating within the visible light spectrum, has emerged as a promising candidate for next-generation indoor wireless systems. Leveraging existing lighting infrastructure, Li-Fi offers vast unlicensed bandwidth, inherent security through confined coverage, and high potential data rates. However, despite these advantages, Li-Fi faces significant deployment challenges—most notably limited coverage, susceptibility to line-of-sight (LOS) blockages, sensitivity to device orientation, and degraded performance in high mobility scenarios. Furthermore, existing MAC-Iayer protocols in Li-Fi often borrow from Wi-Fi standards, such as CSMA/CA, which are not fully optimized for the unique characteristics of visible light communication (VLC) networks. The initial stage of this research addressed MAC-Iayer inefficiencies in heterogeneous Li-Fi environments. A hybrid CSMA/CA—HCCA uplink MAC protocol was developed to dynamically switch between contention-based and contention-free modes depending on device types and network load, significantly improving throughput, delay, and collision probability in diverse traffic scenarios [1]. While this approach enhanced medium access efficiency, it did not fully address performance degradation caused by user mobility and dynamic channel conditions. To tackle mobility-induced challenges, an Orientation-Aware Multi-AP Li-Fi Network (OAM-LiFiNet) was proposed [2]. This framework leveraged real-time SINR measurements and channel metrics to dynamically adjust device orientation, thereby mitigating interference and improving throughput under user movement. Although effective, Li-Fi's dependence on LOS links meant that the system remained vulnerable to blockage events caused by static obstacles or transient movement within the environment. Recognizing the importance of blockage modeling, the research introduced FixOM and SAM [3], two novel approaches for quantifying the impact of obstacles in Li-Fi environments. FixOM modeled stationary obstructions using geometric analysis, while SAM incorporated both complete and partial shadowing effects for a more realistic performance representation. These models provided valuable insights for Li-Fi deployment planning but also highlighted that blockage mitigation within a pure Li-Fi framework could not entirely eliminate service interruptions. This realization motivated the transition toward hybrid Li-Fi/Wi-Fi networks (HLWNs), which combine Li-Fi's high-speed links with Wi-Fi's broader coverage and robustness. Prototype testbeds were developed [4, 5] to evaluate hybrid systems in realistic indoor scenarios, demonstrating superior throughput, handover performance, and service continuity compared to standalone technologies. While hybrid networks improved coverage and throughput, modern MU-MIMO Wi-Fi still suffered from one significant drawback—the high overhead of channel state information (CSI) feedback. This feedback is essential for spatial multiplexing but consumes substantial wireless resources, reducing spectral efficiency and limiting achievable throughput, especially in dense multi-user scenarios. In this thesis, this disadvantage is addressed by utilizing LiFi links to carry CSI feedback through the proposed WiLiConnect and WiLiConnect-Opt systems [6, 7]. By offloading CSI transmission to Li-Fi access points, Wi-Fi capacity is freed for user data, drastically reducing overhead and substantially improving sum-rate performance in hybrid deployments. Building on this foundation, the research advanced to link aggregation and mobilityaware resource allocation strategies for HLWNs. The final stage of the research focused on advanced link aggregation algorithms—LA-SINR, LA-EQ0S, and FLADA [8, maximized combined Li-Fi/Wi-Fi throughput while meeting QoS and fairness requirements. These were complemented by a mobility-aware handover optimization strategy [10] that used a two-stage approach: offline linear programming for static users and a search-space pruning mechanism with re-optimization for mobile users near AP borders. Analytical modeling of outage probabilities [11] in LA-enabled HLWNs further validated their superior robustness compared to standalone networks. Through these interconnected contributions, this thesis presents a holistic, multi-layered framework for high-capacity, mobility-friendly indoor networks. By starting from MAC-Iayer improvements in standalone Li-Fi, addressing mobility and blockage, overcoming Wi-Fi's CSI feedback bottleneck via Li-Fi, integrating hybrid Li-Fi/WiFi architectures, and culminating in advanced link aggregation and handover optimization, the research provides a comprehensive roadmap for deploying next-generation indoor wireless systems that meet the evolving demands of smart homes, offices, and industrial environments.