2.2 The units of capacity

This report expresses the channel capacity in bits per second because ultimately the system designer wants to know how much data can be pumped through the channel how quickly using the given power available. The laser properties, optics efficiency, pointing accuracy, and space and atmospheric losses all affect $C$, but only through their influence on $\bar n_s$, $\bar n_b$, $T_s$, and $T_d$. Hence we express the capacity as a function of the following parameters:

$$C = C(M,\bar n_s,\bar n_b,T_s,T_d,$$detector$$).$$

The units in which $C$ is expressed affect the parameter values which maximize $C$. This fact, which might seem surprising at first, implies that work on maximizing photon efficiency (e.g., [McE81,BKL82,Les83,Ham98a,Ham99]) does not necessarily help determine the maximum data rate possible on the channel.