The problem is stated as follows.
\begin{equation*} \underset{\|\mathbf{x}\|_2 \le \gamma}{\max}\quad\frac{1+|\mathbf{y}_1^H\mathbf{x}|^2}{1+|\mathbf{y}_2^H\mathbf{x}|^2}, \end{equation*}
where $\gamma > 0$ and $\mathbf{y}_1$, $\mathbf{y}_2 \in \mathbb C^n$ are given, and $\mathbf{x} \in \mathbb C^n$ is the optimization variable.
Hint :
You can try writing $\bf y_2 = \bf a + \bf b$ with $\bf a=\frac{\langle \bf y_1, \bf y_2 \rangle}{\lVert \bf y_1\rVert^2} \bf y_1$ and then observe that $|\langle \bf y_2, \bf x\rangle|^2 = |\langle \bf a, \bf x\rangle|^2+|\langle \bf b, \bf x\rangle|^2$. You want the second term to be $0$ hence $\bf x$ and $\bf y_1$ to be collinear. Then you can write your problem as \begin{align*} \max_{x:\,|x|\cdot\lVert \bf y_1\rVert \leq \gamma} \frac{1+|x|^2 \cdot \lVert \bf y_1 \rVert^2}{1+|x|^2\cdot|\bf y_2^H\bf y_1|^2} \end{align*}