Vehicular visible light communications

Abstract

To enable high-speed wireless communications, higher frequencies in the radio frequency (RF) spectrum are required to satisfy the increased demand for communication capacity. Even though researchers and industry have improved the cellular network architecture to increase the data rate, as well as decrease the latency and achieve a better quality of services, most of the RF spectrum has been allocated for specific use. Therefore, it is expected to find an alternative to expand the spectrum. The emergence of visible light communication (VLC) provides a promising alternative to RF-based communications. Light-emitting diodes (LEDs) have been adopted in many illumination applications, due to the significant improvement in solid-state lighting technology. The investigation of VLC began in close-distance indoor scenarios and, over time, was adapted to wider applications.

When it comes to transportation, in Intelligent Transportation System (ITS), VLC is also a powerful candidate. The current LED headlights can be connected with a modulator module and therefore, are possible to transmit signals as a transmitter. The structure of the vehicle will remain unchanged and the reuse of headlights, to some extent, saves energy. Therefore, it is possible to integrate VLC into vehicles to allow vehicular VLC (V-VLC) in vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) schemes. However, the signal-to-noise ratio (SNR) performance of VLC in outdoor scenarios is limited for many reasons, such as adverse weather conditions, artificial light and solar irradiance.

The transmitter itself, the headlight, has a narrow irradiation angle and is able to show a higher directionality, which also limits the coverage and the data rate.

In terms of the analytical channel models in vehicular communication (VC), most of the existing models are static and the same as the indoor scenarios.

However, a moving vehicle changes its orientation and distances of transceivers, which leads to a time-varying communication channel. This time-related channel shows uncertainty, diversity and dependency. However, most of the current research lacks the analysis of this time-related channel. An analytical method has been presented in this thesis to calculate the time-related channel DC gain in dynamic VC systems. This method relates the road roughness to the change in the orientation of the transceivers by using the mechanical movement models and therefore, is able to give a more realistic result, with regards to a real road measurement. Moreover, other vehicle-related variables also influence the SNR on the rough road, which is novel in the work.

In addition to the dynamic channel models, the low sensitivity of the receiver is another challenge for outdoor environments. A single-photon avalanche diode (SPAD) was applied to receive the signals, which enabled a longer communication distance. However, the non-linearity of the SPAD and its unique mechanism limited the performance and therefore, the SNR expression for a specific modulation method should be derived differently and expressed explicitly. In this thesis, the mechanism of the SPAD was analyzed and the SNR expression in single-input single-output (SISO) and orthogonal frequency division multiplexing (OFDM)-based vehicular communications was derived, in terms of the SPAD parameters and positions of transceivers.

From the SISO perspective, the results showed many limitations and therefore, multi-input multi-output (MIMO) scenarios were introduced to subsequently improve the coverage. Due to the existing non-linearity issue and the uniqueness of the mechanism in the SPAD array, the channel gain cannot be obtained directly. In this part of the work, the SNR in MIMO scenarios was expressed. In terms of the number of transceivers, some typical scenarios were considered. In addition, a general 2 × 2 spatial multiplexing (SM)-MIMO was also studied. Further performance improvements can be achieved by employing a longer sampling duration and reducing the spatial correlation.