Influence of Fluorescent Lights

To build systems as robust and efficient as possible it is necessary to consider further channel impacting factors. Fluorescent lights are very common in office buildings making them a not negligible quantity. Multiple scientists for example in [6] and [7] have adressed this matter by researching the impact of fluorescent lights on the wireless channel. As a result it was discovered that fluorescent lights introduce a fast fading process in the scale of double the powerline frequency to the channel. This is due to the reflection and absorption of the signal by the ionised gas. The alternating current changes the dielectric properties of the gas rapidly which leads to signal variations in small time intervals.

The 802.11n channel model [3] includes this effect in the office scenarios as fluorescent lights are mainly prevalent in those. It is introduced through an amplitude modulation of 3 paths and can mathematically be described by $$c'(t) = c(t)\left[\alpha\left(\sum\limits_{l=0}^{2}A_l \exp{(j(4\pi(2l+1)f_m t+ \phi_l)}\right)\right]$$ where

• $c'(t)$ is the modulated path value
• $c(t)$ is the original path value
• $A_l$ a parameter for the amplitudes [0, -15, -20]dB
• $l$ denotes an index
• $f_m$ denotes the powerline frequency (50 Hz for Germany/60 Hz for the US)
• $t$ denotes the time
• $\phi_l$ a statistical independent random number uniformly distributed over $[0, 2\pi)$
$\alpha$ is a normalization constant that is calculated by $$\frac{IC_{\mathrm{rand}}}{IC_{\mathrm{real}}}$$ where $IC_{\mathrm{real}}$ denotes the ratio of the modulated path energy to the original one and $IC_{\mathrm{rand}}$ is defined by $$IC\mathrm{rand} = X^2$$ where $X$ is a statistical independent gaussian distributed random variable with variance $\sigma^2=0.0107^2$ and standard deviation $\mu=0.0203$ [3].