Efficient Service Differentiation and Energy Management in Hybrid WiFi/LiFi Networks

Asim Ihsan, Muhammad Asif, Hossein Safi, Iman Tavakkolnia, Harald Haas · 2025

Read Efficient Service Differentiation and Energy Management in Hybrid WiFi/LiFi Networks by Asim Ihsan, Muhammad Asif, Hossein Safi, Iman Tavakkolnia, Harald Haas (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-16061.txt.

Introduction

Asim Ihsan and co-authors run hybrid WiFi and LiFi networks with an eye on both service tiers and energy bills, in a 2025 paper. LiFi beams data on light, which adds capacity indoors but ties signals to lamps. The team's predict-and-optimize plan sorts each user to the right slice of service in real time.

A deep model, trained on past performance data at a slow control layer, guides the fast layer's slice picks. Math routines then tune the beam settings for the best energy score under delay caps. LED limits, rate floors, and power budgets all enter the sums. Simulations test the plan's balance of speed and thrift. Dense indoor networks gain a route to tiered service without wasting power.

Catalog record

CREATORAsim Ihsan, Muhammad Asif, Hossein Safi, Iman Tavakkolnia, Harald Haas
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE67 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-16061

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Efficient Service Differentiation and Energy Management in Hybrid WiFi/LiFi Networks Asim Ihsan, Muhammad Asif, Hossein Safi, Iman Tavakkolnia, Harald Haas Abstract In this paper, we propose an innovative predict-and-optimize algorithm designed for hybrid WiFi/LiFi networks, aiming to achieve service differentiation while maximizing energy efficiency (EE). The proposed framework utilizes multi-access technology real-time intelligent controller (mATRIC) to dynamically predict the appropriate network slice for each user based on historically monitored key performance indicators (KPIs). This prediction is facilitated by a deep learning model trained using the resilient backpropagation algorithm, with training conducted on KPIs data at the universal non-real time RAN intelligent controller (non-RT RIC). This trained model enables real-time slice selection by mATRIC. In the subsequent phase, the algorithm focuses on optimizing EE of hybrid network as a function of precoding vectors for the predicted slices by employing techniques from sequential convex approximation and the inner approximation method. We introduce novel approximations to convert non-convex objective functions and constraints into convex forms, and develop an iterative algorithm to achieve sub-optimal solutions. Additionally, the EE maximization problem, ensures alignment with end-to-end latency requirements. It also addresses the various constraints inherent to hybrid systems, such as input signal limitations for LiFi LEDs, data rate restrictions, and power budget considerations. Simulation results validate the effectiveness of the proposed algorithm, demonstrating significant improvements in EE while ensuring service differentiation within hybrid network environments. I Introduction and motivation The exponential growth in mobile data traffic and the diverse requirements of emerging applications necessitates innovative solutions in wireless communication systems. The advent of 5G technology and subsequent generations has highlighted the potential of network slicing. Network slicing allows for the differentiation of services by segmenting the network into distinct slices, each adapted to specific application needs [ 1 ] . This capability supports the heterogeneous needs of enhanced mobile broadband (eMBB), ultra-reliable low-latency communication (URLLC), and massive machine-type communications (mMTC), while striving to balance performance with energy efficiency. Despite its advantages in managing diverse service requirements, network slicing poses challenges, particularly related to energy consumption and operational costs, raising concerns about sustainability [ 2 ] . As network operators aim to deliver a broad range of services, integrating energy-efficient strategies becomes essential. Open radio access network (O-RAN) is an initiative to develop an open, disaggregated RAN architecture that promotes interoperability and flexibility by standardizing interfaces and components. This approach allows for greater innovation and cost-efficiency, enabling seamless integration of different radio technologies into a cohesive network ecosystem. O-RAN will unify various radio technologies, including non-3GPP technologies, Wi-Fi and LiFi, to satisfy the diverse requirements of future wireless networks and create a more integrated and versatile communication infrastructure [ 3 ] . For control and optimization of RAN functions across multi-access technologies, including all wired and wireless technologies, the realizing enabling architectures and solutions for open networks (REASON) project plays a crucial role. REASON is dedicated to advancing the O-RAN initiative by developing multi-access technology real-time intelligent controller (mATRIC), an innovative framework designed to manage and optimize a diverse array of access technologies [ 4 ] . It acts as a central component, receiving policy directives from the universal non-real time RAN intelligent controller (non-RT RIC) and executing real-time control and optimization of multi-access technologies. By utilizing collected data, mATRIC ensures dynamic network performance and adaptability and supports various xApps. Here, xApps are microservice based application that are executed on the mATRIC to enhance network functionality and efficiency in real-time. In O-RAN, non-3GPP access networks, like WiFi and LiFi, can be integrated using non-3GPP interworking functions. Both WiFi and LiFi networks can connect through the same interworking function, but each network uses its own separate IP subnetwork. To fully capitalize on the benefits of mATRIC for hybrid WiFi/LiFi network, it is crucial to extend network slicing to accommodate these hybrid technologies and applications within the RAN ecosystem [ 5 ] . Consequently, supporting network slicing for each technology is essential. For Wi-Fi and Li-Fi networks, slicing can be effectively applied to address their unique requirements, allowing for the creation of tailored virtual slices that optimize performance and resource allocation. This approach ensures that each network can meet specific needs, such as high-density Wi-Fi environments or high-speed Li-Fi communications, enhancing overall efficiency and service quality [ 7 , 8 , 6 ] . By applying network slicing to a hybrid setup, it is possible to optimize the use of both WiFi and LiFi resources in a coordinated manner through mATRIC. mATRIC can dynamically allocate resources among different access technologies and network slices, ensuring optimal performance for eMBB, URLLC, and mMTC applications. To effectively support the diverse service requirements of eMBB, URLLC, mMTC in a hybrid WiFi/LiFi network, it is crucial to consider the specific transmission needs of each service type. Each of these services has distinct performance and reliability requirements that impact the design of the network and the optimization strategies employed. For URLLC, which demands extremely low latency and high reliability, the transmission duration plays a critical role. Given that the end-to-end latency requirement for URLLC is approximately 1 millisecond [ 9 , 10 ] , a transmission duration of 0.05 milliseconds [ 11 , 12 ] , is sufficient to meet this stringent requirement [ 13 ] . Finite block length transmission is crucial here due to the short block lengths involved, which affect error rates. Traditional Shannon capacity, applicable to longer block lengths, does not accurately reflect achievable rates for URLLC. Instead, the Polyanskiy bound [ 14 ] , considering finite block lengths, provides a more accurate measure of achievable rates, which are lower than those predicted by Shannon capacity. This necessitates algorithms that handle finite block lengths effectively to ensure high reliability and meet URLLC’s stringent performance requirements. For mMTC, which manages numerous devices with intermittent communication needs, finite block length transmission is particularly advantageous. While mMTC’s latency requirements are less stringent than those of URLLC, short block lengths are still valuable due to the nature of mMTC’s communication. This technology often involves transmitting short payloads, where ensuring reliability and efficiency is crucial. Finite block length transmission enables the system to handle these brief communication intervals effectively, maintaining reliable performance, thus aligning well with the demands of mMTC [ 15 ] . Dealing with the finite block length constraints for URLLC and mMTC, alongside traditional large block length transmission for eMBB, presents a complex challenge for resource allocation. This scenario introduces novel optimization problems that require advanced strategies for efficient resource management [ 16 ] . In hybrid WiFi/LiFi networks, most studies have contributed towards key issues such as EE [ 17 , 18 ] , load balancing [ 19 , 20 ] , handover [ 21 ] , user association [ 22 , 23 ] , and access point selection [ 24 ] . However, a notable gap exists in integrating service differentiation with EE, an aspect that remains unexplored in hybrid WiFi/LiFi networks. Addressing this gap, our work focuses on developing strategies that not only balance performance and energy consumption but also meet diverse service requirements of different application, thereby fully exploiting the potential of hybrid networks. Our major contribution is summarized as follows • A novel predict-and-optimize algorithm is introduced for

Read the complete text