Given i.i.d. Rayleigh fading, known (at BS) large scale-fading parameters and pilot based channel estimation for a Massive MIMO system, can we have a better channel estimator than the one achieved with the MMSE estimation?
I mean, for the above mentioned case, would there be a better channel estimator?
Thanks and Kind Regards,
Felipe
Please see partial update procedures to attain better estimates. A second approach could be ML sequence estimation based on Soft-input-soft-output detection..
Best Wishes
No, not if you consider the MSE as your performance criterion and only use the received pilot for channel estimation.
But one can also take the received data into account in a data-aided channel estimator.
Thanks Dr. Emil. Assuming MSE is my performance criterion and that I only use pilot based channel estimation, how could I prove that there is no better estimator?
Is it only because it has the minimum mean squared error among all the estimators?
Kind regards,
Felipe
If MSE is your performance criterion, the MMSE approach is the optimum solution for construction!
Ivan is right, the MMSE estimation is by definition the one that minimizes the MSE.
Dear respected colleagues,
It may be useful if you read the paper cited in the following link about the sphere detector for spatial spatial multiplexing MIMO: Article FPGA Implementation of Sphere Detector for Spatial Multiplex...
Best wishes
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude