Studies of optical wireless communications: random orientation model, modulation, and hybrid WiFi and LiFi networks
Purwita, Ardimas Andi
Cisco has predicted that by 2022, as a result of the emerging internet-of-things traffic, there will be an internet protocol (IP) traffic explosion. An increasing amount of radio frequency (RF) spectra has been allocated to accommodate this mobile data surge, such as the recently allocated sub 6 GHz band for WiFi. A further prediction is that by the year 2035, all RF spectra will be fully utilized. This means there is a need for additional spectrum, such as the optical spectrum. One vision of future wireless networks is to aggregate multiple wireless access technologies. For example, optical wireless communications (OWC) that have narrow coverage, but a very high area spectral efficiency, can potentially complement RF communications to provide a higher peak data rate. A hybrid WiFi and light fidelity (LiFi) network is one of these examples. Compared to other OWC technologies, such as visible light communications or optical camera communications, LiFi supports bidirectional and multi-user communications. These features are also the main features of WiFi, and WiFi is predicted to carry the majority of global mobile data traffic in the future. Based on this prediction, the primary question in this thesis is how much LiFi supports WiFi in efficiently handling mobile data traffic in hybrid WiFi and LiFi networks. This is answered by calculating an offloading efficiency, which is the ratio of data that is transferred over LiFi compared to the total data. Even with the current advancements in LiFi, a few intermediate studies are needed. • The first contribution of this thesis is to model randomly-oriented mobile devices. A random orientation is needed in order to capture users’ behavior while they move and operate mobile devices. In addition, a random blockage model must be investigated as LiFi signals can be blocked by opaque objects. The main purpose of considering the random orientation and random blockage models is to ensure that the offloading efficiency is evaluated by using realistic assumptions. • The second contribution of this thesis fall under studies of single-carrier and multi-carrier modulations. The conventional pulse amplitude modulation and single-carrier with frequency domain equalization (PAM-SCFDE) is improved by adding non-linear filters or index modulation. In the low-to-moderate spectral efficiency region, up to 3 dB gain can be achieved by means of these improvements. In the high spectral efficiency region, an orthogonal frequency division multiplexing (OFDM)-based system, which is based on the common mode of the physical (PHY) layer of the ongoing LiFi standardization, i.e., IEEE 802.11bb, is used. By exploiting the fact that the wireless optical channel has a low-pass filter characteristic, an in-phase and quadrature wavelength division multiplexing (IQ-WDM) system that uses the PHY of IEEE 802.11bb is proposed. Up to 2 dB gain can be obtained by IQ-WDM compared to the common mode PHY of IEEE 802.11bb. • For the final contribution of this thesis, the OFDM system from IEEE 802.11bb is abstracted to emulate large-scale networks and to calculate the offloading efficiency. A real transport control protocol (TCP)/IP stack, which has multipath TCP (MPTCP) to support multi-connectivity between WiFi and LiFi networks, is also deployed to emulate real traffic. By calculating the offloading efficiency with such methodology over many channel realizations considering the proposed random orientation and blockage models, average offloading efficiencies of 64.54% and 75.85% are obtained for residential and enterprise scenarios, respectively. These results show the significant potential of OWC to complement RF communications.