The end goal for performing a link analysis is to ensure that the received power at the ground station is strong enough to be detected by the transceiver board, and to be differentiated from the noise in the receiver's environment. Additionally, the selected Modulation Techniques have different bit error rates for different signal-to-noise ratios (or more specifically, for different Eb/N0 ratios).
In link analysis, there are two main ways of performing this signal-to-noise analysis.
The SNR method of link analysis uses the following formula to calculate a signal-to-noise power ratio:
$$ SNR=\frac{P_R}{N}=\frac{P_TG_TG_RL}{kT_NB} $$
Here, $N=kT_NB$ is the system noise power (see Noise Temperature). Notice that $P_R$ is calculated using the Friis Transmission Equation. As a quick reminder, $P_R$ is the received power, $k$ is the Boltzmann constant, $T_N$ is the receiver noise temperature, and $B$ is the bandwidth.
The SNR value calculated from this equation is then compared (subtracted if using dB, divided if using pure values) with the required SNR for a particular modulation technique to obtain a "system link margin" - a value which must be greater than 0 dB in order for the link to work.
The Eb/N0 method is similar to the SNR method, but substitutes bandwidth within the equation for bitrate, $R$:
$$ \frac{E_b}{N_0}=\frac{P_TG_TG_RL}{kT_NR} $$
This equation determines the ratio of the energy per bit to the spectral noise density (noise power per Hz of bandwidth).
The information from this page is taken from The New SMAD, chapter 21.1, as well as "Eb/N0 Explained" by Jim Pearce.