NP: the values of the vector □ are chosen from well-known options, it mean every value is expected to be a value out of 4 or 8 or 16 ….etc. So can we first train the deep learning model to produce x_p and then x_d can be estimated using the trained model? It is like joint estimation of both vectors. Where □_p are the well-known values which are a part of the total vectorx. The below figure is the structure of my diagram: I was wondering, If I know the first N/4 values of □, can I use deep neural network or CNN to recover the other values of □ ? Which model might be the best to be the best in that situation? The signal y is wise-point multiplied with random vector H which is representing the channel Y=H.y I have a signal vector □ whose length is N with complex values, it is multiplied with a well-known real unitary matrix D with size N x N to yield a vector y = □D. I am little bit newbie in deep neural network, but I need to use it to solve an issue I met. Gze Asks: Selecting the deep learning model for jointly estimation and detection
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |