HS-PSK: High-rate transfer over RoI(s)

HS-PSK is an operating mode being standardized at IEEE 802.15.7m. This post describes what it is for and how it operates.

Figure 1 illustrates an example in which an Rx can select the link properly for an intended usage. The link setup procedure is solved by the RoI-Signaling waveform, while this high-rate waveform is additionally transferred if the link is set.

RoI---Section 5--vehicular.png
Figure 1 – Example of high-rate optical transfer usage over a selected RoI-link

You can see a presentation of our measurement for Optical SNR in V2X model in the following video. This practical measurement is helpful to later evaluations.


Technical Considerations for HS-PSK Waveform

The dimmable waveform is required for high-rate data stream modulation. The clock of dimming control for this high-rate communication is synchronized to the clock of the low-rate RoI signaling stream. Such a selection of a modulation scheme for high-rate stream must provide acceptable performance under dimming. At least, it must reach reliable performance at two dimming levels, a darker dimming level, and a brighter dimming level, for the integration of the low-rate stream via its dimming.

Dimmable feature is the mandatory requirement. Moreover, the second technical consideration of the required sampling rate also plays an important role in the proposed system. Particularly, a high-frame-rate camera is to receive a high-rate modulated waveform, and thus the sampling rate determined by the camera frame rate must follow the Nyquist sampling theorem. The frame rate requirement is not about technical concern (because Mfps camera is available on the market) but about the economic (cost effective) expense. As discussed in the IEEE 802.15.7m, Intel introduced tens of kilo-frame-per-second (kfps) system which is feasible for RoI camera design with acceptable complexity. Consequently, the high-rate waveform is designed being decodable by the no-more-than-20 kfps camera.

To leverage the data rate without raising the frame rate requirement, multiple LEDs array is considered. The challenging issue really comes when the dimming requirement and multi-LEDs are jointly set.

What’s new in HS-PSK Waveform

By newly designing the waveform for high-rate data transfer over the selectable RoI link, novel contributions of the waveform can be highlighted as follows.

(1) High-resolution dimming:

It is mandatory to have 2-dimming levels to generate the binary states for the RoI signaling waveform. However, more dimming levels options are still a desire for high-quality light emission. This newly designed waveform can support dimming in steps of 12.5% without declining the communication performance.

(2) Lower frame rate requirement with intelligent decoding ability:

Undeniably, a camera cannot decode data if the state of LED is fuzzy (later called the x_state). The designed waveform is able to support the decoding with the fuzzy state of LEDs. The fuzzy decoding concept can save significantly the frame rate requirement of the camera (fuzzy decoder has at least twice lower frame rate requirement than the binary-based decoder).


HS-PSK Operation

HS-PSK is a hybrid waveform of S2-PSK (a low-rate RoI Signaling Waveform, operating at a low optical clock rate e.g., 10Hz) and DS8-PSK (a high-rate mode by using multiple times higher optical clock rate, e.g., 10kHz, a dimmable waveform).

Fig 8.png
Figure 2 – Proposed OCC-based V2V system using car taillights and cameras

A reference architecture to implement HS-PSK is as in Figure 3. At first, two data streams (a low-rate stream and a high-rate stream) are modulated individually in which a RoI singling scheme (e.g., S2-PSK) modulates a low rate data stream and a dimmable signaling scheme (e.g., DS8-PSK) modulates a high rate data stream. The outputs of the S2-PSK modulator are to control the dimming levels of the outputs of DS8-PSK; therefore, the output waveforms are hybrid modulation of S2-PSK and DS8-PSK.

1 (D4-Fig 1854 -HS-PSK ----reference architecture.png
Figure 3 – Functional Architecture for HS-PSKEnter a caption

Figure 4 illustrates the output waveform of HS-PSK. The rearmost outputs to drive LEDs are twinkle waveform of the high-rate DS8-PSK and the low-rate S2-PSK.

3- (D4-FIg 186)  HS-PSK waveforms.png
Figure 4 – HS-PSK waveform to modulate vehicular light sources

HS-PSK Decoding

Because there are two simultaneous links are available on the hybrid waveform, the demodulation of these two data streams can be performed with a dual-camera receiver system as follows.

  • A lower frame rate camera (e., the frame rate should be greater than the S2-PSK optical clock rate) is to detect the S2-PSK signal. This camera configures its exposure time as slow enough to filter higher tones out of received signal. Such the integration time of sampling allows only the frequency lower than the cut-off frequency that is determined by the shutter speed going through.

In the system, either a global shutter or a rolling shutter camera can be used for S2-PSK.

  • A high-speed camera (e., the frame rate should be greater than the DS8-PSK optical clock rate) is to decode the high-rate DS8-PSK signal.

In the system, a global shutter camera is recommended to sample the high-rate light intensity change of the groups of LEDs.

HS-PSK Decoding Example

There are two cases: (1) camera makes sampling for clear ON/OFF states, (2) camera makes sampling in the transition time of the pulses, so there are fuzzy states here. Fortunately, our decoder can successfully decode both cases without any need for extra redundancy.

J.2. Fig --DS8-PSK -----HS-PSK decoder.png
Figure 5 – DS8-PSK decoding procedure (without fuzzy states)
J.2. Fig --DS8-PSK -----HS-PSK decoder---x_state.png
Figure 6 – DS8-PSK decoding procedure (presence of fuzzy states)


IEEE 802.15.7m ongoing standard has provided the encoding and decoding tables for both cases. Standard is coming soon by the end of this year.



The generation of hybrid-waveform dealing with the high-rate data transfer requirement along with its reliable link setup and tracking need is introduced. The integration of two data streams is implemented over dimming.

HS-PSK is designed to support fuzzy-state decoding so that the frame rate of camera requirement is just a half of typical decoder. Furthermore, this HS-PSK encoder enables a Neural-Network support for the decoder. We are implementing AI-assisted decoder and our current results show that its performance is better than a typical one.






Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s