Why to use machine learning in communication networks?
Answers
Answered by
0
Wireless communications study in academia is predominantly model based. You will encounter a significant degree of analytical derivations based on probabilistic models - such as predicting a transmitted symbol in the presence of additive thermal noise and channel fading both of which are well characterized probabilistically (AWGN, Rayleigh, respectively - for instance) - this could be MAP estimation or MMSE. Overall, you could completely get away with data processing and simulations in academic study of wireless communications.
The story is different in communications industry, though. Channel, noise and interference models are not as well-defined in reality. This is where your ML study could come handy. Here you could use ML techniques along with data collected by your company to come up with models, or skip the modeling step and directly build numerical optimization techniques.
The story is different in communications industry, though. Channel, noise and interference models are not as well-defined in reality. This is where your ML study could come handy. Here you could use ML techniques along with data collected by your company to come up with models, or skip the modeling step and directly build numerical optimization techniques.
Similar questions