I have the following specific polynomial of degree 4: \begin{align} P(x) &= (x_1^2 + x_2^2)^2 + a(x_1 + i x_2)^2(x_1^2 + x_2^2) + \overline{a} (x_1-ix_2)^2(x_1^2 + x_2^2)\\ &+b(x_1+ix_2)^4 + \overline{b}(x_1-ix_2)^4, \, x=(x_1,x_2)\in \mathbb{R}^2. \end{align} In addition, I know that coefficients $a,b$ are such that \begin{equation} P(x) \geq c(x_1^2+x^2_2)^2 , \, c > 0. \end{equation}
I have a conjecture that $P$ must have only simple (complex) roots. For example, if one will look for the similar polynomial of degree 2, i.e.: \begin{align} &Q(x) = (x_1^2 + x_2^2) + a(x_1 + i x_2)^2 + \overline{a} (x_1-ix_2)^2,\\ &Q(x) \geq c(x_1^2 + x_2^2) \end{align} then one can directly show that the corresponding roots will be only simple. And condition of positivity is used to have that the determinant is not zero.
Sorry to all. Question is not at all well posed. By simple characterstics I meant that in the decomposition of the polynomial it won't appear any monomial of the form: \begin{equation} (x_1 + z x_2)^k,\, z \in \mathbb{C}, \, k\geq 2. \end{equation} The case when $a = 0, b = 0$ is the obvious counterexample, so the conjecture is absolutely not true.