It is an interesting paper by the way.
I think "the reference" refers to reference 25 which appears in superscript immediately before! Wertz, James R., 2011a. “Atmospheric Drag and Satellite Decay,” Sec. 9.4.4 in Space Mission Engineering: The New SMAD, ed. by J. Wertz, D. Everett, and J. Puschell, Hawthorne, CA: Microcosm Press, 2011.
However, in the original SMAD it looks like Eq. 6.21 through 6.23 already give what you are asking about:
$$a_D = -\frac{1}{2} \rho (C_D A/m) V^2 \ \ \ \ \ \text{(6-21)}$$
We can approximate the changes in semimajor axis and eccentricity per revolution, and the lifetime of a satellite, using the following equations:
$$\Delta a_{rev} = -2 \pi (C_D A/m)a^2 \rho_p \exp(-c)(I_0 + 2 e I_1) \ \ \ \ \ \text{(6-22)}$$
$$\Delta e_{rev} = -2 \pi (C_D A/m) a \rho_p \exp(-c) (I_1 + e(I_0 + I_2 )/2) \ \ \ \ \ \text{(6-23)}$$
where $\rho_p$ is atmospheric density at perigee, $c=ae/H$, $H$ is density scale height (see column 25, Inside Rear Cover), and $I_i$ are Modified Bessel Functions of order $i$ and argument $c$. We model the term $m/(C_D A)$, or ballistic coefficient as a constant for most satellites, although it can vary by a factor of 10 depending on the satellite’s orientation (See Table 8-3).
This gives you $\Delta a_{rev}$ and $\Delta e_{rev}$, the change in $a$ and $e$ per revolution, respectively, but it would still up to you to choose a value for $m/(C_D A)$ or ballistic coefficient and to try to represent the atmosphere's variable density profile with a local Scale Height approximation necessary for this analytical approximation to work, rather than a proper numerical model based on a realistic density profile).
below: From your linked paper https://en.wikipedia.org/wiki/Scale_height Wertz et al. SSC12-IV-6, 26th Annual AIAA/USU Conference on Small Satellites. There are some straight sections on these log-lin plots, where a scale height approximation would not be too unreasonable.