Hi! I think that gene (and miRNA) expression is too much a cell type-specific phenomenon to look for a "general" control cell line. I can suggest you to match at least the cell type of your cancerous cell line if you do not have the same cell line in a non-cancerous state (something like before they acquired the cancer-inducing mutation). It will be better to compare more than one control cell line with more than one cancerous cell line since there could be some individual variations also... What is you cell line model? I'm not a cancer expert anyway so my answer could be a bit naive... Anyway best of luck with your experiments!
I do not recommend this, unless you have no other cell-lines.
miRNA expression varies between cell types, and differentiation states. Therefore, the most ideal control would be the untransformed primary cell of the cancer cell type.
I do not have as much experience working with miRNAs, but there are 2 people that I would recommend contacting or reading their work: