Source-linked AI summary
Impact of Multipath Reflections on the Performance of Indoor Visible Light Positioning Systems
Wenjun Gu, Mohammadreza Aminikashani, Mohsen Kavehrad
TL;DR
Indoor VLC positioning can use LED transmitters and photodiode RSS measurements, but prior analyses often neglected reflections from room surfaces. This paper models those effects with a CDMMC impulse-response method, quantifies positioning errors across a room, and evaluates calibration approaches. Multipath reflections considerably reduce accuracy, especially in outer regions, while signal selection improves the reported RMS error.
Problem
The paper addresses limited positioning analysis that considers line-of-sight channels without accounting for multipath reflections in indoor VLC environments.
Method
The system uses photodiode RSS measurements for distance estimation and applies CDMMC to calculate the optical-channel impulse response in a room with reflecting surfaces.
Results
Multipath reflections considerably decrease positioning accuracy, especially in outer regions; total-room RMS error is 0.5589 m with reflections versus 0.0040 m without them.
Takeaways & Limitations
Selecting only the strongest LED signals reduces total RMS error to 0.3240 m when four signals are used.
Abstract
from arXiv · showhide
Visible light communication (VLC) using light-emitting-diodes (LEDs) has been a popular research area recently. VLC can provide a practical solution for indoor positioning. In this paper, the impact of multipath reflections on indoor VLC positioning is investigated, considering a complex indoor environment with walls, floor and ceiling. For the proposed positioning system, an LED bulb is the transmitter and a photo-diode (PD) is the receiver to detect received signal strength (RSS) information. Combined deterministic and modified Monte Carlo (CDMMC) method is applied to compute the impulse response of the optical channel. Since power attenuation is applied to calculate the distance between the transmitter and receiver, the received power from each reflection order is analyzed. Finally, the positioning errors are estimated for all the locations over the room and compared with the previous works where no reflections considered. Three calibration approaches are proposed to decrease the effect of multipath reflections.
I. INTRODUCTION
Indoor VLC positioning is motivated by GPS limitations indoors and uses LED transmitters with photodiode-based RSS measurements. The paper focuses on the neglected effect of multipath reflections in realistic rooms and studies a combined impulse-response method.
- GPS signals attenuate severely through solid walls, motivating alternative indoor-positioning technologies.
- VLC supports indoor positioning because illumination infrastructure can also provide communication and positioning functionality.
- The proposed positioning approach detects RSS with a photodiode, converts signal power to transmitter distance, and applies lateration to estimate receiver coordinates.
- Prior positioning analyses generally considered only line-of-sight channels, although wide-beam LEDs and finite-FOV receivers also capture reflections from room surfaces.
- The paper investigates multipath-induced positioning distortion using a combined deterministic and modified Monte Carlo impulse-response approach.
B. Channel Access Method
The system separates LED transmissions with synchronized time slots and uses a combined deterministic and modified Monte Carlo procedure to compute optical-channel impulse responses. Surface elements and recursively generated reflected rays contribute to the total received power.
- B. Channel Access Method: Time division multiplexing assigns synchronized, nonoverlapping frame slots to LED transmitters while others provide constant illumination.
- C. Impulse Response Analysis Method: The CDMMC method computes first-reflection contributions deterministically and higher-order reflections with modified Monte Carlo sampling.
- C. Impulse Response Analysis Method: Room surfaces are divided into small square elements, and the photodiode and elements are treated as receivers for received-power calculations.
- C. Impulse Response Analysis Method: Each surface element becomes a point source whose power is determined using the surface reflection coefficient.
- C. Impulse Response Analysis Method: The channel impulse response sums the line-of-sight contribution with recursively generated contributions from successive reflections.
III. POSITIONING ALGORITHM
The positioning algorithm derives received signal strength from the optical impulse response and uses it to estimate transmitter–receiver distance. The model incorporates transmitter power, propagation geometry, and receiver optical parameters.
- Received signal strength for transmitter i is obtained after computing the channel impulse response.
- The transmitted power is assumed to be 5 W for logic 1 and 3 W for logic 0.
- The random-ray coordinate system uses α for the angle from the z-axis, β for the projected azimuth, and m for the Lambertian order.
- The received-power relation includes transmitter power, distance, irradiance and incident angles, optical-filter transmittance, concentrator gain, and detector area.
- The transmitter and receiver parameters used by the algorithm are listed in Table II.
B. Least Square Estimation
The receiver coordinates are estimated from transmitter coordinates and measured distances by converting the positioning equations into matrix form and applying least-squares estimation.
- Receiver coordinates (x, y) are related to the detected transmitters’ coordinates through a group of positioning equations.
- Subtracting the last three equations from the first produces a linearized system for estimating the receiver position.
- The linearized equations are expressed in matrix format before solving for the coordinates.
- The least-square estimator minimizes the squared error to obtain the estimated receiver coordinates ˆX = [ˆx ˆy]T.
IV. SIMULATION AND RESULT
The simulation evaluates multipath effects at three representative receiver locations with different expected reflection severity in a room containing 16 ceiling-mounted LED bulbs.
- Three locations represent severe, medium, and weak multipath conditions: corner A, wall-adjacent edge B, and central point C.A is (0, 0, 1.2), B is (4, 0, 1.2), and C is (4, 4, 1.2).
- The system model assumes 16 LED bulbs installed on the ceiling.
- The inner region is bounded by dashed lines, while the remaining area is designated as the outer region.
A. Impulse Response Analysis
Impulse-response analysis shows that multipath reflections can be comparable to the line-of-sight component at some locations but negligible at others. Reflection strength therefore varies across the room and affects expected positioning accuracy.
- At Location A, reflected impulse-response amplitudes are comparable to LOS, causing large positioning errors.
- At Location B, reflection amplitudes decrease significantly compared with Location A, so positioning accuracy is expected to improve.
- At Location C, the LOS component almost dominates the total impulse response and reflections are negligible.
- The received power from each reflection order is analyzed because RSS is used to estimate transmitter–receiver distance.
B. Power Intensity Distribution Analysis
Received power from each LED transmitter is examined across reflection orders because RSS-based distance estimation directly affects positioning performance.
- Reflection components can be comparable to LOS power for several LED signals, affecting distance estimation based only on direct attenuation.At the corner point, only the first LED has LOS power much greater than reflections; the other five signals have comparable reflection components.
C. Analysis of Positioning Accuracy
Multipath reflections substantially worsen positioning accuracy compared with the ideal no-reflection case, with errors concentrated in outer and corner regions. Across the room, the RMS error rises from 0.0040 m without reflections to 0.5589 m with them.
- Without reflections, positioning errors are low throughout the room and mostly within 0.005 m.The no-reflection case includes only small thermal and shot-noise errors.
- With reflections, corner and edge locations are especially affected, while positioning accuracy remains satisfactory near the room centre.Reflections are described as weak at the centre and severe in outer areas.
- Some reflected-power positioning errors reach 1.7 m, although most errors remain below 1 m.
- 0.5589 m RMS error occurs across the entire room with reflections, versus 0.0040 m when reflections are neglected.The outer and inner regions have RMS errors of 0.8173 m and 0.2024 m, respectively, in the reflection case.
V. DISCUSSION AND CALIBRATION APPROACHES
The paper proposes calibration approaches to address the substantial positioning degradation caused by multipath reflections. Nonlinear least-square estimation directly optimizes the original objective and reduces errors relative to linear estimation.
- V. DISCUSSION AND CALIBRATION APPROACHES: Three calibration approaches are proposed because multipath reflections considerably affect positioning accuracy, especially in the outer region.
- A. Nonlinear estimation: Nonlinear least-square estimation directly minimizes the original positioning objective instead of relying on the linear least-square solution.The approach uses an iterative trust-region-reflective algorithm initialized with the linear estimate.
- A. Nonlinear estimation: Most nonlinear-estimation errors are within 0.8 m, with only a few exceeding 1 m and a worst error around 1.5 m.
- A. Nonlinear estimation: 0.6871 m outer-region RMS error is achieved with nonlinear estimation, compared with the linear method; inner-region and entire-room RMS errors become 0.1401 m and 0.4642 m.
B. Selection of LED Signals
The section evaluates LED-signal selection and denser bulb layouts as ways to reduce positioning errors caused by multipath reflections. Selecting strong signals lowers RMS error, while a 1.5 m layout improves performance overall but can make linear initialization fail at some locations.
- Signal selection: 0.3240 m total RMS error is obtained when the four strongest LED signals are selected for coordinate estimation.The corresponding RMS errors are 0.4046 m and 0.3527 m when six and five strongest signals are selected, respectively.
- Signal selection: 0.4828 m outer-region RMS error versus 0.0849 m inner-region RMS error shows that four-signal selection benefits the outer region more strongly.The four strongest signals are used with nonlinear estimation in the reported best results.
- Denser LED layout: With 25 LED bulbs spaced 1.5 m apart, nonlinear-estimation RMS error is 0.3121 without selection, 0.2922 m with six signals, and 0.2699 m with five signals.The denser layout produces a more uniform light-intensity distribution, supporting improved positioning accuracy.
- Denser LED layout: Four-signal selection can prevent coordinate recovery by linear estimation at some locations because three selected bulbs may lie in the same row, creating a singular Matrix A.In those cases, the average detected LED-bulb coordinates initializes nonlinear estimation, but this is less accurate than the linear estimate and increases positioning error.
- Denser LED layout: Most positioning errors in the best five-signal nonlinear-estimation scenario fall within 0.5 m, although some large errors remain at room edges and corners.The inner area is described as satisfactory, while edge and corner locations remain more problematic.
VI. CONCLUSIONS
The paper investigates indoor visible-light positioning while accounting for multipath reflections in a typical room. It compares positioning performance with reflection effects, proposes three calibration approaches, and reports improvements from nonlinear estimation, LED-signal selection, and dense bulb layouts.
- Findings: Multipath reflections considerably decrease positioning accuracy, especially in the outer region of the room.The study compares positioning errors with earlier analyses that do not account for reflections.
- Calibration approaches: Three calibration approaches are proposed to improve positioning performance under multipath reflections.The approaches include nonlinear estimation, LED-signal selection, and a dense LED-bulb layout.
- Calibration approaches: Nonlinear estimation decreases positioning error, while LED-signal selection removes signals severely affected by multipath reflections.The conclusion also links dense bulb layouts with a more uniform light-intensity distribution.
- Calibration approaches: A dense LED-bulb layout results in a more uniform light-intensity distribution and improved positioning performance.