PaperlayerView exact revision on arXiv

Source-linked AI summary

Hybrid LiFi and WiFi Networks: A Survey

Xiping Wu, Mohammad Dehghani Soltani, Lai Zhou, Majid Safari, Harald Haas

arXiv:2001.04840v1cs.IT

TL;DR

Indoor wireless traffic growth and RF spectrum constraints motivate combining LiFi’s high-speed links with WiFi’s broad coverage. This survey presents an HLWNet framework, reviews performance and achievements, and examines challenges and applications. It identifies handover, load balancing, interference management, and centralized control as central integration concerns.

  • Problem

    Indoor wireless networks must accommodate growing traffic despite RF spectrum limitations and the short coverage and blockage sensitivity of LiFi.

  • Method

    The paper surveys HLWNet frameworks, deployment and access schemes, modulation, illumination, backhauling, performance metrics, challenges, and applications.

  • Results

    The survey reports that HLWNets improve network performance and support applications including indoor positioning and physical layer security.

  • Takeaways & Limitations

    Efficient HLWNet integration requires attention to handovers, low-complexity load balancing, interference management, and centralized control such as SDN.

Abstract

from arXiv · show

To tackle the rapidly growing number of mobile devices and their expanding demands for Internet services, network convergence is envisaged to integrate different technology domains. A recently proposed and promising approach to indoor wireless communications is integrating light fidelity (LiFi) and wireless fidelity (WiFi), namely a hybrid LiFi and WiFi network (HLWNet). This type of network combines the high-speed data transmission of LiFi and the ubiquitous coverage of WiFi. In this paper, we present a survey-style introduction to HLWNets, starting with a framework including the network structure, cell deployment, multiple access schemes, modulation techniques, illumination requirements and backhauling. Then, key performance metrics and recent achievements are reviewed. Further, the unique challenges faced by HLWNets are elaborated in many research directions, including user behavior modeling, interference management, handover and load balancing. Finally, we discuss the potential of HLWNets in application areas such as indoor positioning and physical layer security.

I. INTRODUCTION

HLWNets combine LiFi’s high-speed transmission with WiFi’s broad coverage to address indoor wireless demands and spectrum limitations. The survey reviews their framework, performance, challenges, and application potential.

  • LiFi exploits the approximately 300 THz optical spectrum and can provide peak data rates above 10 Gbps using a single LED.It also offers licence-free spectrum, operation in RF-restricted areas, and security benefits, but typically covers only a few meters per AP.
  • HLWNets integrate LiFi and WiFi, combining LiFi’s high-speed transmission with WiFi’s ubiquitous coverage.The hybrid concept was introduced for indoor wireless communications and is reported to outperform standalone LiFi or RF systems.
  • Prior HLWNet studies considered system design and resource allocation but focused on stationary users, omitted comprehensive overviews, or lacked frequency-reuse interference management.These gaps motivate a survey covering the state of the art, unique challenges, open issues, and research directions.
  • The survey addresses HLWNet system design, resource allocation, user mobility, interference management, handover, and load balancing.It also reviews illumination requirements, backhauling, modulation, multiple access, performance metrics, and application services.
  • HLWNets can improve throughput and latency while supporting indoor positioning and physical layer security.These application benefits are presented alongside the network’s communications role.

A. Network Structure

HLWNet structure involves centralized management, indoor cell deployment choices, and multiple-access schemes adapted from WiFi and proposed for LiFi. Deployment geometry affects SINR coverage and practicality, while CSMA/CA coordinates channel access through sensing and optional RTS/CTS handshaking.

  • A. Network Structure: An SDN-based hypervisor can manage all LiFi and WiFi APs by decoupling control and data planes for flexible association and resource allocation.An SDN-enabled switch connects the APs and supports cross-layer resource allocation.
  • Cell Deployment: Hexagonal, lattice, and Poisson point process deployments model different indoor AP arrangements, with lattice deployment most practical for ceiling-mounted LiFi lamps.Hexagonal deployment provides an upper-bound analysis, while PPP is uncommon because illumination requirements are difficult to satisfy.
  • Cell Deployment: Lattice deployment has SINR coverage probability very close to hexagonal deployment, whereas PPP provides the worst SINR coverage probability among the three.The comparison concerns the noted deployment models for LiFi networks.
  • Multiple Access: HLWNet multiple access includes WiFi’s standardized CSMA/CA and candidate LiFi schemes such as TDMA and OFDMA.LiFi standardization remains in progress through IEEE P802.15.13, ITU G.vlc, and IEEE 802.11bb.
  • 1) CSMA/CA:: CSMA/CA senses the channel before transmission, uses random binary exponential backoff, and can employ RTS/CTS four-way handshaking to reduce collisions.After CTS reception, the node transmits data and receives an ACK from the serving AP.

2) TDMA:

The section compares TDMA, OFDMA, and NOMA as HLWNet access schemes, then contrasts single- and multiple-carrier modulation for LiFi. It highlights synchronization, interference, QoS allocation, and spectral-efficiency trade-offs.

  • TDMA: TDMA assigns users separate time slots across the available modulation bandwidth, enabling sequential transmissions.TDMA can be directly used in HLWNets, particularly for downlink.
  • TDMA: TDMA requires synchronization and is more difficult for uplink mobile UEs because randomly located terminals have different propagation times.Timing advance can compensate random delays, while inter-channel interference can degrade cell-edge performance.
  • OFDMA: OFDMA allocates different subsets of subcarriers to users, allowing individual data rates to be controlled according to QoS.It is considered for downlink transmission in both WiFi-6 and LiFi HLWNs.
  • NOMA: NOMA lets users share the same time and frequency resources through power-domain multiplexing and successive interference cancellation.Users with good channels receive lower transmit power, while users with poor channels receive higher power.
  • Modulation techniques: Multiple-carrier modulation offers higher spectral efficiency and data rates than single-carrier schemes, which are more vulnerable to inter-symbol interference as rates increase.Single-carrier methods remain attractive for low-speed or low-medium-rate applications because of their simpler implementation.
  • Modulation techniques: LiFi requires optical OFDM variants because intensity modulation and direct detection need positive, real-valued signals.DCO-OFDM and ACO-OFDM are identified as two unipolar optical-OFDM solutions.

1) DCO-OFDM:

DCO-OFDM converts QAM-modulated subcarrier data into a real, positive waveform suitable for LiFi intensity modulation and direct detection.

  • DCO-OFDM: DCO-OFDM begins by mapping information bits to QAM symbols and grouping K/2 − 1 consecutive symbols into an OFDM frame.The frame is then processed by the IFFT module.
  • DCO-OFDM: Hermitian symmetry is applied to the OFDM frame so that the IFFT output becomes real-valued in the time domain.The subcarrier symbols are constrained using conjugate-symmetry conditions.
  • DCO-OFDM: A cyclic prefix is added after the IFFT to combat inter-symbol interference caused by the dispersive wireless channel.The samples are then converted from digital to analog before optical transmission.
  • DCO-OFDM: A DC bias is added to the analog waveform to ensure a positive signal for optical intensity modulation.The resulting current signal drives the LED to generate the optical signal.
  • DCO-OFDM: When η = 3, less than 1% of the signal is clipped, making clipping noise negligible.η is the conversion factor used in the clipping condition.

2) ACO-OFDM:

ACO-OFDM creates a unipolar LiFi waveform without adding a DC bias by using only odd subcarriers, trading spectral efficiency for energy efficiency and clipping robustness.

  • ACO-OFDM: ACO-OFDM uses only odd subcarriers to carry information, which reduces spectral efficiency.Its OFDM frame places zeros on the unused subcarriers and satisfies Hermitian symmetry.
  • ACO-OFDM: Compared to DCO-OFDM, ACO-OFDM sacrifices half of the spectrum to obtain a unipolar time-domain signal.The resulting IFFT signal is real-valued and anti-symmetric.
  • ACO-OFDM: ACO-OFDM’s anti-symmetry ensures that zero-level clipping does not lose information data.This property supports unipolar optical transmission without a DC bias.
  • ACO-OFDM: For a given constellation, ACO-OFDM has half the BER of DCO-OFDM because data symbols occupy only half of the subcarriers.The paper compares the schemes across spectrum and energy efficiency, data rate, hardware complexity, and multipath robustness.

III. PERFORMANCE METRICS

HLWNet studies use coverage, fairness, and related metrics to evaluate service quality and resource allocation. The surveyed literature reports that hybrid LiFi/WiFi systems can improve fairness through load balancing and user reassignment.

  • A. Coverage Probability: Coverage probability measures the likelihood that a user's received SINR exceeds a specified threshold.In HLWNets, it depends on both user location and device orientation.
  • A. Coverage Probability: Coverage probability can represent the fraction of users, time, or cellular area whose SINR exceeds the threshold.
  • B. Fairness: Hybrid RF/LiFi networks can enhance fairness by transferring users with poor QoS from overloaded access points to less-loaded ones.Reported approaches include dynamic load balancing, joint power allocation, and fuzzy-logic load balancing.
  • B. Fairness: The grade of fairness is higher when the throughput share of each technology is close to its share of connected users.

C. Spectral Efficiency

The survey reports that hybrid LiFi/RF networks improve spectral, area-spectral, and energy efficiency relative to standalone systems. These gains are associated with spatial bandwidth reuse, joint resource allocation, and load balancing.

  • C. Spectral Efficiency: Hybrid RF/VLC networks provide greater spectral efficiency than standalone RF or VLC networks.Joint AP assignment and resource allocation can improve spectral efficiency twice compared with conventional signal-strength selection.
  • D. Area Spectral Efficiency: Area spectral efficiency compares system data rate with indoor area and captures spectral performance for a given user density.
  • D. Area Spectral Efficiency: LiFi can achieve several hundred times higher area spectral efficiency than a femtocell because of high bandwidth reuse.
  • D. Area Spectral Efficiency: A hybrid LiFi/femtocell RF system increases average area spectral efficiency by at least two orders of magnitude over standalone RF.
  • E. Energy Efficiency: Hybrid RF/VLC networks are reported to be more energy efficient than standalone RF or VLC networks.Studies use joint bandwidth, power allocation, and load balancing to optimize energy efficiency or achievable data rate.

F. Network Throughput

HLWNet throughput is shaped by user mobility, device orientation, and blockage, while hybrid LiFi/WiFi deployments generally improve throughput over standalone systems.

  • Hybrid LiFi/RF networking can halve outage probability at a target throughput of 30 Mbps compared with standalone LiFi.Outage data rate is the probability that achieved data rate is below threshold R_th.
  • Hybrid networks improve user throughput by more than 30 Mbps over standalone LiFi and more than 40 Mbps over standalone RF systems.The comparison evaluates field-of-view effects on average user throughput.
  • Load-balancing mechanisms improve average user throughput across regular and merged cell formations and can serve mobile users through WiFi while quasi-static users use LiFi.Merged cells combine neighboring cells to support mobility and reduce inter-cell interference.
  • User behavior modeling: User throughput and QoS depend on mobility, random device orientation, and light-path blockage, motivating mobility models such as random waypoint.The random waypoint model is widely used for indoor and outdoor mobility simulation.
  • User behavior modeling: Device orientation is especially important in LiFi because its short-wavelength optical channel is sensitive to receiver rotation.Most prior LiFi studies neglected random device orientation because suitable models were lacking.

C. Light-path Blockage

Light-path blockage is a distinctive LiFi vulnerability in HLWNets, but WiFi coverage and receiver diversity provide ways to preserve connectivity and improve performance.

  • Opaque objects such as human bodies can block LiFi links, whereas WiFi is comparatively robust to blockage.When blockage occurs, increasing LiFi transmission power cannot compensate for the resulting error rate.
  • When a LiFi link is blocked, users can connect to WiFi, yielding average downlink rate R̄ = P_bR_wifi + (1 − P_b)R_LiFi.P_b is the blockage probability, while R_LiFi and R_wifi denote LiFi and WiFi data rates.
  • Hybrid WiFi/LiFi networks improve area data rate under blocker densities and can use evolutionary-game load balancing under blockage and inter-cell interference.The reported scheme allocates LiFi and WiFi resources while considering shadowing.
  • At low blockage density, blockers can reduce neighboring-cell interference, so blockage is not always destructive to user data rate.
  • Research directions: Angular, multidirectional, and omnidirectional receivers improve robustness to blockage and random orientation through spatial diversity.These receiver structures can also offload some traffic from WiFi access points.

V. INTERFERENCE MANAGEMENT

HLWNets require interference management tailored to LiFi’s optical constraints, combining cancellation, avoidance, receiver design, and cross-technology coordination.

  • LiFi and WiFi operate in different spectra, so their integration avoids direct cross-technology interference while LiFi still faces intra-network interference.
  • Interference cancellation: LiFi precoding must accommodate optical-signal non-negativity, commonly by modifying conventional schemes with a DC biasing vector.Examples include zero forcing, block diagonalization, and dirty paper coding.
  • Blind interference alignment: Blind interference alignment reduces interference without exact transmitter CSI by pairing time-selective and frequency-selective users, but only to some extent.
  • Blind interference alignment: Blind interference alignment outperforms TDMA only above 50 dBm optical transmit power in one cited study, a level described as impractical for illumination LEDs.
  • Power control: Power control in LiFi must preserve illumination requirements, and without successive interference cancellation it cannot function as an interference-avoidance method.
  • LiFi-specific techniques: Angle diversity receivers reduce interference using multiple narrow-FoV photodiodes, while orthogonal schemes such as TDMA, OFDMA, and SDMA avoid interference at transmission.

1) Frequency Reuse:

Frequency reuse can manage inter-cell interference in LiFi, but mobility, limited LED bandwidth, wavelength separation, and cross-technology coordination remain important constraints.

  • Frequency Reuse: Frequency reuse assigns neighboring LiFi cells regular frequency patterns to avoid inter-cell interference; fractional reuse divides cells into regions.Strict and soft frequency reuse are identified as fractional-reuse variants.
  • Frequency Reuse: Frequency/time hopping lowers the probability that users share a time-frequency block but requires strict transmitter-receiver synchronization.
  • Challenges and Research Directions: User mobility complicates LiFi interference management because each LiFi access point covers a relatively small area, while frequency reuse implementation remains open.Commercial LEDs typically provide 10–20 MHz modulation bandwidth, limiting single-wavelength reuse.
  • Challenges and Research Directions: Using multiple wavelengths preserves each color LED’s full bandwidth, but photodiodes generally cannot separate wavelengths without optical filtering or spectrum-sensor arrays.Adaptive filters and spectrum-sensor arrays are proposed solutions, with array area proportional to supported wavelengths.
  • Challenges and Research Directions: HLWNets can shift severe LiFi interferers to WiFi, but network-interactive interference management remains a research gap.
  • Handover: Small LiFi coverage and changing receiver orientation produce frequent handovers, making handover overhead a critical throughput factor.A single LiFi access point may cover only 2–3 meters in diameter.
  • Handover: Optimal LiFi coverage areas are reported as 2–8 m^2 depending on user density and handover overhead, while handover rate peaks at device tilts of 60°–80°.

B. Vertical Handover (VHO)

Vertical handover (VHO) connects users from LiFi to WiFi when LiFi connectivity is lost, but its higher processing time and WiFi’s lower capacity make handover selection and load balancing critical.

  • VHO usually transfers users from LiFi to WiFi, whereas HHO changes the serving AP within the same access technology.
  • LiFi connectivity can be lost when opaque objects block the light path or when the receiver orientation deviates from the line-of-sight path.
  • VHO takes longer than HHO because the air interface changes, and excessive WiFi users can substantially reduce throughput.
  • Not every user experiencing LiFi loss should switch to WiFi, particularly during transient light-path blockage, making HHO/VHO choice a load-balancing problem.
  • Existing handover studies mostly assume single-AP association, while multiple-AP association could maintain connectivity during blockage and avoid handovers.
  • Load balancing: HLWNet load balancing is essential because LiFi and WiFi coverage overlap completely, while WiFi has broader coverage but lower capacity and is therefore prone to overload.
  • Load balancing: Stationary-channel and optimization-based load-balancing methods face mobility constraints because rapidly varying CSI limits their practicable processing time.

B. Mobility-aware Load Balancing

Mobility-aware load balancing addresses changing user associations by incorporating movement and handover costs, but practical deployment remains constrained by response-time and modeling challenges.

  • Mobility-aware load balancing: Stationary load-balancing solutions can repeatedly transfer users between LiFi and WiFi as users move across LiFi APs.
  • Mobility-aware load balancing: A mobility-aware approach uses achievable data rate and moving direction for user preferences, while AP preferences depend on the sum data rate of served users.
  • Mobility-aware load balancing: User mobility affects load balancing through handover cost and response time, which includes algorithm processing and control-unit/AP information exchange.
  • Mobility-aware load balancing: Existing mobility-aware load-balancing methods account for handover cost but have been evaluated only offline through simulations.
  • Research challenges: Arbitrary receiver orientations can change LiFi coverage areas, while light-path blockages require dynamic load balancing.
  • Research challenges: Current studies generally use absent or fixed data-rate requirements, although practical user requirements vary over time and can produce slot-level traffic imbalance.
  • Applications: HLWNet application research covers indoor positioning and physical layer security, alongside broader wireless-network challenges.
  • Applications: Indoor-positioning methods are classified mathematically as triangulation, proximity, or fingerprinting, and by information type as RSS, TOA, TDOA, or AOA.

2) IPS in Hybrid Networks:

Hybrid LiFi/WiFi networks extend indoor positioning and security applications by combining LiFi’s precision and security properties with WiFi or RF coverage and redundancy, while important scope gaps remain.

  • Indoor positioning: LiFi positioning errors are 0.1-0.35 m, compared with 1-7 m for WiFi, and LiFi can provide denser beacons through existing lighting infrastructure.
  • Indoor positioning: Hybrid positioning systems combine LiFi and RF technologies through staged localization, with reported errors below 130 cm, within 20 cm, and 5.8 cm.
  • Indoor positioning: The reported hybrid positioning studies focus on detecting two-dimensional positions.
  • Physical layer security: LiFi’s security advantages include opaque-object blocking, shorter single-AP range, and LoS-concentrated transmission power usually above 85%.
  • Physical layer security: HLWNet security studies combine LiFi resilience with WiFi redundancy, while also examining secrecy, energy harvesting, and power consumption.
  • Challenges: Open positioning issues include 3-D localization and orientation detection, while physical-layer security requires balancing strong security with convenient implementation.
  • Conclusion: The survey concludes that HLWNets can improve network performance and support indoor positioning and physical-layer security, but efficient integration still requires addressing key challenges.
Loading 2001.04840v1…