Artificial intelligent assistant

How to linearize a constraint of the form of a product? Is there a way to linearize a constraint of the form: $$\prod\limits_{ i=1 }^{ n }y_i\geqslant b,$$ where $y_i$ are discrete variables in the set $\\{1,2,\ldots,2^m\\}$ for some $m>2$ and $b$ is a positive real number.

Consider \begin{equation} \textrm{log} \left( \prod_{i=1}^n y_i \right) = \sum\limits_{i=1}^{n} \textrm{log}(y_i) \geq \textrm{log}(1) = 0 \end{equation}

As $y_i = 2^{k_i}$ then \begin{equation} \textrm{log}(y_i) = \textrm{log}\left(2^{k_i}\right) = k_i \textrm{log}(2) \end{equation}

Therefore an equivalent linearized constraint is: \begin{equation} \sum\limits_{i=1}^{n} k_i \geq 0 \end{equation}

But as $k_i \in \mathbb{N}$ the constraint is always valid.

xcX3v84RxoQ-4GxG32940ukFUIEgYdPy b14272136893a7d1bddc537617b943a3