I have to choose between a transaction journal on Network science and engineering with impact factor ( 2.5) and IEEE internet of things journal with 9 impact. I know that people prefer transactions journals. What do you recommend me to do ? Which one is more better in my CV as a Ph.D. student.
I don't know about your specific journals, but Impact Factor is a fake measure and should be ignored. Both the idea of measuring "impact" by number of citations and the methodology itself to collect such data are flawed, hence impact factors constitute meaningless information in most cases.
What is important are the norms in your scientific community. Ask your colleagues what is better. Do not look at "impact factors".