⌘OWIN6G EXPERIMENT LABSENSORS / OPTICS / NETWORK
OPTICAL WIRELESS SENSOR NETWORKS FOR 6G

From vehicle sensors to the digital twin.

Follow sensor samples through serial, optical and Ethernet communication.

Simulated environment
01 Vehicle → roadside infrastructure
V2I OPTICAL UPLINK
Sensors → vehicle → optical transmitter → photodetector → receiver → gateway
PROXIMITYClear
ULTRASOUND— m
LiDAR— m
FRAMES SAMPLED0
ETHERNET INGRESS0
V2I FRAMES LOST0
Each frame bundles all three sensor samples. Light is converted to an electrical signal, decoded, then encapsulated for Ethernet. Animation timing is illustrative.
02 Optical / Ethernet network
● LIVE
FSO power monitor Light present
DRAG ROUTERS TO ARRANGE · SELECT ANY DEVICE TO INSPECT
TelemetryUDP floodPoisoned dataEthernetFSO / V2I lightFibreManagement
00:00
Network telemetry delivered0 / 0 sentSince reset · after RSU ingress
Resolved delivery ratio—Delivered ÷ resolved telemetry
Network telemetry lost0 packets0 greyhole · 0 congestion · 0 FSO
Mean network transit—RSU gateway → Server · simulated
04 LiDAR truth vs. digital twin
Waiting for telemetry
Physical LiDAR24.0 m
Twin reading—
Difference—
Last 60 simulated secondsWaiting for first packet
Telemetry collector→Neo4j state→DT orchestrator→Ryu
Queued packets 0 actual 0 reportedSensor → Server mean: —
05 Node inspector
ROUTER
R2Forwarding telemetry

Intermediate Ethernet router on the selected route.

Queue occupancy0 / 36
Forwarded0
Dropped0

Model: 18 packets/s per router, 36-packet queues, 4 sensor frames/s. Each logical network link takes 0.65s for visibility; drawn distance is not propagation time.

06 Event stream
Simulation time
    OWIN6G FIELD GUIDE

    Follow the same data across layers.

    The vehicle samples proximity, ultrasound and LiDAR. Each telemetry frame bundles the three readings, crosses the serial interface and drives an optical transmitter. The roadside photodetector produces an electrical signal. Receiver electronics decode it; the RSU gateway sends a network packet to R1.

    LiDAR is a ranging sensor here. Its sensing light is separate from the optical transmitter carrying the V2I data. Serial framing, modulation, analog circuits and Ethernet encapsulation are represented as stages, not bit-level implementations.

    1. Follow the primary path: R1 → Ethernet → MC1A → fibre → 2 m FSO → fibre → MC1B → Ethernet → R2 → Ethernet → R3 → Server.
    2. Interrupt a beam: V2I interruption prevents new sensor frames from reaching network ingress. FSO interruption loses network frames crossing the primary span. The power monitor indicates beam presence only; it is not a calibrated power measurement.
    3. Select the backup: R1 → MC2A → 3 km SSMF → MC2B → R3 bypasses both FSO and R2. Route changes apply when packets enter the network; existing packets keep their path. Switching is manual.
    4. Inject an attack: UDP flood is injected at the RSU’s network output. Greyhole selectively drops telemetry at a router. Sensor poisoning adds a bias to LiDAR samples before optical transmission; twin poisoning changes the reported LiDAR value and queue state.

    Reading the metrics

    V2I loss is counted before network ingress. Network loss includes greyhole, congestion and FSO loss. The delivery ratio uses resolved network telemetry: received / (received + all network losses). Pending frames are excluded. Mean network transit starts at the RSU gateway; sensor-to-server delay includes the physical stages. All counters are cumulative since Reset.

    Displayed seconds are deliberately slowed model time, not measured Ethernet/optical propagation. Three sensor channels use synthetic readings. The 2 m FSO and 3 km fibre labels come from your diagram; no link budget, BER, atmospheric model, live hardware, Neo4j or Ryu controller is connected. Removing attacks leaves optical interruptions unchanged; Reset restores the full experiment.