Non Line-of-Sight Optical Wireless Communication using Neuromorphic Cameras

TL;DR

This work investigates passive non-line-of-sight optical wireless communication using neuromorphic cameras that detect illumination changes as asynchronous events. It develops an end-to-end system that converts reflections from static objects into detectable events and demodulates them into data using OOK and N-pulse modulation, including an adaptive scheme. It demonstrates that adaptive N-pulse modulation improves data rate and reliability across dark and ambient lighting, with performance strongly dependent on object reflectivity, size, finish, and proximity to the light source. The results illustrate the dual utility of neuromorphic cameras for joint sensing and communication in indoor environments and outline future directions for mobility, synchronization, and error correction.

Abstract

Neuromorphic or event cameras, inspired by biological vision systems, capture changes in illumination with high temporal resolution and efficiency, producing streams of events rather than traditional images. In this paper, we explore the use of neuromorphic cameras for passive optical wireless communication (OWC), leveraging their asynchronous detection of illumination changes to decode data transmitted through reflections of light from objects. We propose a novel system that utilizes neuromorphic cameras for passive visible light communication (VLC), extending the concept to Non Line-of-Sight (NLoS) scenarios through passive reflections from everyday objects. Our experiments demonstrate the feasibility and advantages of using neuromorphic cameras for VLC, characterizing the performance of various modulation schemes, including traditional On-Off Keying (OOK) and advanced N-pulse modulation. We introduce an adaptive N-pulse modulation scheme that dynamically adjusts encoding based on the packet's bit composition, achieving higher data rates and robustness in different scenarios. Our results show that lighter-colored, glossy objects are better for NLoS communication, while larger objects and those with matte finishes experience higher error rates due to multipath reflections.

Figures (11)

• Figure 1: This diagram shows event generation in a neuromorphic camera based on light intensity changes. As the intensity crosses the positive threshold (Bias Diff On, red dashed line), an "on" event (red circle) is triggered. When it crosses the negative threshold (Bias Diff Off, blue dashed line), an "off" event (blue circle) is generated. Each pixel operates asynchronously, producing timestamped events only during changes, resulting in high temporal resolution, minimal data redundancy, and low latency.
• Figure 2: Spectral response of analog filters present in event cameras. The frequency filtered out by the high-pass and low-pass filters can be controlled by adjusting the bias values in the camera. The observed event rate is the band present in the overlap area between the high pass and low pass filters.
• Figure 3: Conceptual illustration of the proposed joint sensing and communication using optical wireless and neuromorphic camera on a robot
• Figure 4: Various objects used in the experiment to test different surface properties and their effects in sensing and communication. From left to right - Glossy white flask (red box), Matte finish measuring tape (yellow box), Compact mirror (blue box), glossy finish leather ball (green box), Matte finish white colored google nest (magenta box), Sound reflecting foam (completely absorptive) (brown box)
• Figure 5: This figure illustrates the overall system block diagram of the receiver and its main tasks that help us achieve this modality of joint sensing and communication. Firstly, we have a neuromorphic camera that is observing the scene. Then from the USB 3.0 interface, we get the events of that scene. Then we do a periodic events accumulation framing to get the event frame representation as shown in the next step, and then we find the object of interest and region to focus on using the RoI refinement technique. Further from there we have our demodulation and decoding logic that converts all the events to bits.