In consideration of spectral efficiency TDD is more spectral efficient as it uses single frequency for operation.

What do you mean with that? You will not use one frequency but an interval of frequencies, and you either split it between uplink and downlink in time or by dividing the bandwidth into two subbands.

I would rather say that the 3 dB loss in SNR that you get in TDD operation will reduce the spectral efficiency as compared to FDD. (http://ma-mimo.ellintech.se/2018/06/22/disadvantages-with-tdd/)

On the other hand, TDD makes it much easier to implement Massive MIMO in order to bring up the spectral efficiency by spatial multiplexing of users.

Hence it depend on the network operator to chose duplex schemes

In theory, yes. But in practice, when you buy a spectrum license, you are (normally) required to use a particular duplex scheme. There are particular bands assigned for TDD and other bands for FDD (where you have two subbands, one for uplink and one for downlink, that are well separated to avoid interference).

Hello,

In Frequency Division Duplexing (FDD) different frequencies are allocated for the uplink and downlink respectively. Whereas in Time Division Duplexing (TDD) the time slot is divided for Uplink and Downlink( Transmission and Reception occurs on same channel).

Duplexer is not required for TDD inverse is for the FDD.So TDD is cost effective.

Regulatory changes are required to change capacity for FDD, whereas Uplink and Downlink capacity can be changed in TDD.

Guard period required to separate uplink and dowlink is large in TDD. Whereas it is defined and smaller in case of FDD.

Latency in case of TDD is more as the time interval for transmission and reception is different, whereas it is same for FDD.

In consideration of spectral efficiency TDD is more spectral efficient as it uses single frequency for operation.

The MAC layer complexity is more in TDD as the time interval between transmission and reception synchronization is to be maintained.

Hence it depend on the network operator to chose duplex schemes

Dear Montaha,

I agree with Sagar and add to him. I will will not repeat which he said.

TDD has a reciprocal channel in both up link and down link as they share the same channel at different time slots. If one assumes that that the channel is stationary at the two time slots the up link and down link then one do not need to estimate the channel for the down link which make great savings for the MIMO systems specially MIMO systems.

FDD has more complex front end because of the uplink transmits on a frequency band different from that of the down link requiring two frequency synthesizers as well as filters. Then more hard ware and more power consumption.

The last comment is that the MAC layer for TDD is more complex than the FDD because of frame synchronization.

There is many literature comparing the two duplexing techniques for example:http://www.rfwireless-world.com/Articles/difference-between-TDD-and-FDD.html

In consideration of spectral efficiency TDD is more spectral efficient as it uses single frequency for operation.

What do you mean with that? You will not use one frequency but an interval of frequencies, and you either split it between uplink and downlink in time or by dividing the bandwidth into two subbands.

I would rather say that the 3 dB loss in SNR that you get in TDD operation will reduce the spectral efficiency as compared to FDD. (http://ma-mimo.ellintech.se/2018/06/22/disadvantages-with-tdd/)

On the other hand, TDD makes it much easier to implement Massive MIMO in order to bring up the spectral efficiency by spatial multiplexing of users.

Hence it depend on the network operator to chose duplex schemes

In theory, yes. But in practice, when you buy a spectrum license, you are (normally) required to use a particular duplex scheme. There are particular bands assigned for TDD and other bands for FDD (where you have two subbands, one for uplink and one for downlink, that are well separated to avoid interference).

Thank you all for your answers.

Emil, it is mainly in the last paragraph in your answer where you really get to the point. If someone could develop it more or give a reference. Is it at the core of the network that the duplex scheme is made? If I want to make a phone call, and need to connect to a BS, how the duplex scheme is chosen? Let us assume there are free bands for FDD and TDD.

Thank you!

It is more historical question rather technical. At the beginning FDD devices arrived to the market, so authorities like FCC, ITU and other national ones give more spectrum to FDD, and a little bit reserved to TDD. Later TDD solutions were done, but nobody wish to make changes, because all spectrum ranges already sold. All new spectrum will be allocated for TDD only. There was long discussion about spectrum re-farming, but seems no results.