The MIMO antenna array is described as NTXNr system with Nt is the number of transmit antennas and Nr is the number of the receive antenna

Fot single user antennas the Nr is restricted physically by the sizeof the user equipment to few antennas. As the operating frequency increases the number Nr can increase as the size of a single antenna is reduced.

For conventional single user MIMO, the maximum number of independent paths is = the minimum of Nr and Nt. So it is advantageous to make Nt=Nr=N

where N can take any value limited only by practical consideration.

For massive MIMo with one user normally Nt>>Nr.

For massive MIMO with multi user with every user has only one antenna , the number of user are Nr then the system will be NtxNr.

In order to affect beam forming for space division it is so that one has to make Nt>>Nr to make Nr beams using Nt antennas. As the beam becomes narrower, one can increase Nr in the cell.

So, there are different configuration and sparking from 256 antennas maximum needs evidence.

Best wishes

It is not necessary to have a number of antennas that is a multiple of two. But it is convenient if one describe the antenna indices with as a few bits as possible. More than 256 antenna can certainly be used. It is just a question of cost and implementation complexity.

You can use the same error correcting codes as in other systems. In particular, the channel hardening phenomenon makes fading channels behave as AWGN channels. Hence, codes that have been optimized for AWGN channels work well in Massive MIMO.

Smartphones can only a fit a few antennas (e.g., 2, 4 or 8). LTE phones typically have one transmit antennas and two receive antennas.

Thank you Muhammed, it is a smart and serious question

Definitely for research purposes we can go for more any number, but the question is: is still with the same performance? what about loss complexity? noise? thermal noise? interference? all these factors should be affected radically by increasing number of microstrip patch antennas.

Regarding number: we are looking forward to have regular, symmetric shape. Therefore, technically the best arrangement is the quadratic. having circular or triangular shape is possible. Objectively, we are looking for collaborative radiation to increase gain..... etc.

I think having antennas in a multiple of two is just for the symmetry convenience. To have multiple antennas, we need to have physical separation among them in order their signals do not get correlated. This separation is generally half the wavelength. More antennas will require huge space of very low wavelength (very very high frequency).

Increasing the number of Transmitter Antennas will Increase the number of RF chains connected to it and this allows the system to consume more power, meanwhile you should also think of precoders design for such increment of antennas number. If the number of antennas at the transmitter side is increased then the complexity of transmitter beamforming method using OSTBC increases. meanwhile the SNR level also gets reduced as we all know practically assuming CSI is always not a good idea. Hardware impairments should also be considered. so there are many drawbacks in increasing the number of antennas though you achieve better throughput and efficiency in terms of spectral but you do face reduction in power to which u should work on implementing new techniques which can be mathematically complicated