I'm trying to model convergence of $x\in \mathbb{R}^d$ which follows the following recurrence:

$$x\leftarrow (1-h)^2 x + h\langle x, h\rangle$$


In my application $x$ and $h$ are non-negative, entries of $h$ follow power-law decay with known constant $p\in (1,2)$ and $h$ is small enough so that continuous approximation $x_t\approx \exp(At)x_0$ holds. I need to know how trajectory of $\|x_t\|_1$ depends on $p$ in the case of $t<d$ and $d\to\infty$

Things are easy if we didn't have the the $h\langle x, h\rangle$ term (mixing term). $A$ is diagonal, and approximating $\|x_t\|_1$ with an integral I get formulas which match observed behavior very well.


[![enter image description here][1]][1]

However, keeping the mixing term I get $x_t\approx \exp(At)x_0$ where $A$ is diagonal + rank1 matrix, and this becomes much harder to handle. Applying integration as [before](https://math.stackexchange.com/a/4668173/998) does not produce insights.

Any advice on the approaches to follow to get a nice upper bound on $\|x_t\|_1$ in terms of $p$?

[Notebook](https://www.wolframcloud.com/obj/yaroslavvb/nn-linear/mathoverflow-rank1-recurrence-simple.nb)



  [1]: https://i.sstatic.net/3a3jd.png