4.4.1.2 Normalization of the Stationary Distribution Function

The stationary distribution function $ f_{s}(\vec{k})$ must be normalized as a probability, $ 0<f_{s}(\vec{k})<1$ to guarantee the correct rejection of scattering events. The $ \vec{k}$ space is divided into sub-domains $ \Omega$ of size $ V_{\Omega}=(\Delta k)^{3}$. In the following $ \bar{f}_{\Omega}$ stands for the average distribution function in $ \Omega$ for a given valley and $ n$ is the contribution to the electron density from the same valley. In each sub-domain the electron density is

$\displaystyle n_{\Omega}=\frac{1}{4\pi^{3}}\int_{\Omega}f_{s}(\vec{k})\,d\vec{k}.$ (4.57)

and the average distribution function is given as:

$\displaystyle \bar{f}_{\Omega}=\frac{\int_{\Omega}f_{s}(\vec{k})\,d\vec{k}}{V_{\Omega}}=\frac{4\pi^{3}n_{\Omega}}{V_{\Omega}}.$ (4.58)

Using the before-scattering estimation for the statistical average

$\displaystyle \langle\langle A\rangle\rangle = C\frac{1}{N}\sum_{b}\frac{A(\vec{k}_{b})}{\lambda(\vec{k}_{b})},$ (4.59)

where $ N$ is the number of electron free-flights and the normalization constant $ C$ is given as

$\displaystyle C=\frac{4\pi^{3}N\cdot n}{\sum_{b}\frac{1}{\lambda(\vec{k}_b)}},$ (4.60)

one obtains for $ n_{\Omega}$:

$\displaystyle n_{\Omega}=\frac{1}{4\pi^{3}}\int\Theta_{\Omega}(\vec{k})f_{s}(\v...
...ta_{\Omega}(\vec{k}_{b})/\lambda(\vec{k}_{b})}{\sum_{b}1/\lambda(\vec{k}_{b})},$ (4.61)

where the indicator function $ \Theta_{\Omega}(\vec{k})$ of sub-domain $ \Omega$ has been introduced. Substituting (4.62) into (4.59) the average distribution function is finally obtained:

$\displaystyle \bar{f}_{\Omega}=\frac{4\pi^{3}n}{V_{\Omega}}\cdot\frac{\sum_{b}\Theta_{\Omega}(\vec{k}_{b})/\lambda(\vec{k}_{b})}{\sum_{b}1/\lambda(\vec{k}_{b})}.$ (4.62)

S. Smirnov: