Hello all,
So we meet again.
As you know, the initial focus on 5G NR(as of rel-15) has been on increasing the data throughput (enhanced mobile broadband or eMBB). There are different ways to achieve these high data rate requirements. One is obviously the increased spectrum. The more spectrum you have, the more data you can transmit. That's why mm-wave is important for 5G. Another method is to use rich digital modulation techniques. For example, 64-QAM can be used to send 6 bits in one symbol while 256-QAM can be used to send 9 bits in one symbol. As we move to higher digital modulation techniques, we can send more data per symbol (of course, this is not as easy as it sounds. Higher modulation techniques increase the UE complexity. It will also result in an increased bit error rate if the symbol is not estimated correctly at the receiver). Then there are frequency reuse concepts like MIMO, beamforming, etc.
To achieve the eMBB requirements, RF concepts like massive MIMO & beamforming are stressed more in 5G. I would like to discuss these concepts but I thought it's better to first discuss the basic MIMO before we move to 5G specific features.
MIMO
MIMO is short for multiple in, multiple out. The "In" and "out" always refer to the transmission channel. MIMO is a method for increasing the capacity of a radio link using multiple transmission and receiving antennas by exploiting multipath propagation. MIMO is fundamentally different from smart antenna techniques developed to enhance the performance of a single data signal, such as beamforming and diversity.
If someone asks the question, "How communication happens between a transmitter and a receiver?", one will typically say something like below that there is a transmitter with antenna sending the signal. The signal transmits through the channel and there is a receiver that the same frequency, receives and then decodes the signal.
Suppose, consider a case where there are multiple antennas at the transmitter(say M) and multiple antennas at the receiver(say N). Each antenna at the transmitter is used to send different data blocks on the same spectrum. Will the receiver will be able to decode it? One could say that it may not. Since both antennas are using the same bandwidth spectrum, the interference can be high so that the receiver won't be able to decode the signals. But in MIMO, at the expense of additional hardware and computation complexity, we can increase the number of parallel data streams with the same available bandwidth. That means, with MIMO, we can actually increase the channel capacity.
A key feature of MIMO systems is the ability to turn multipath propagation, traditionally a pitfall of wireless transmission, into a benefit for the user. In such a MIMO system, there are M x N signal paths from the transmit antennas to the receive antennas, and the signals on these paths are not identical. The signal component at each antenna at the receiver includes the signal component from all transmit antennas. Since the MxN multi-paths are different, the channel fading seen at each of the receiver antenna too is different. The signal at the receiver can be unscrambled only when it has good transmit diversity (due to man-made objects or natural obstacles). This diversity in the arrived signal makes it possible for the DSP to unscramble the signal, decode the transmitted data blocks and then makes it into a parallel stream. In some cases, the receiver can send the channel estimation info to the transmitter to adjust the transmission. Such kind of system is called a closed-loop MIMO system.
We should then ask the question - for a MIMO system with M transmit antennas and N receive antennas, how many such parallel data streams are possible? the answer is clear - it is the minimum of (M, N). So, at maximum, under ideal conditions (with a rich multipath environment), it is possible to have min(M, N) times the channel capacity of the SISO system.
Is it always possible to have channel capacity multiplied by min(M, N)? The answer is NO. For MIMO to work, it should have multiple data streams that are uncorrelated in nature. In the case of the MxN system, if there exists LOS between transmitter and receiver, then the no of orthogonal data streams that can be transmitted will come down to 1 as the paths between tx and rx antennas are almost same i.e. the correlation is high. Basically, as the multipath between Tx & Rx decreases, the no of uncorrelated data streams also decreases. The beauty of MIMO is that we are using the naturally available multipath condition(which is considered as a pitfall) to our advantage.
I tried to avoid mathematics to explain the MIMO concept. If you really want to know how exactly it works with mathematics, Please refer to the below IEEE document & chapter 5 in the below book.
Reference: From theory to practice: an overview of MIMO space-time coded wireless systems
Book: https://books.google.co.in/books?id=G5C5ii8O_y0C&printsec=frontcover#v=onepage&q&f=false
So we meet again.
As you know, the initial focus on 5G NR(as of rel-15) has been on increasing the data throughput (enhanced mobile broadband or eMBB). There are different ways to achieve these high data rate requirements. One is obviously the increased spectrum. The more spectrum you have, the more data you can transmit. That's why mm-wave is important for 5G. Another method is to use rich digital modulation techniques. For example, 64-QAM can be used to send 6 bits in one symbol while 256-QAM can be used to send 9 bits in one symbol. As we move to higher digital modulation techniques, we can send more data per symbol (of course, this is not as easy as it sounds. Higher modulation techniques increase the UE complexity. It will also result in an increased bit error rate if the symbol is not estimated correctly at the receiver). Then there are frequency reuse concepts like MIMO, beamforming, etc.
To achieve the eMBB requirements, RF concepts like massive MIMO & beamforming are stressed more in 5G. I would like to discuss these concepts but I thought it's better to first discuss the basic MIMO before we move to 5G specific features.
MIMO
MIMO is short for multiple in, multiple out. The "In" and "out" always refer to the transmission channel. MIMO is a method for increasing the capacity of a radio link using multiple transmission and receiving antennas by exploiting multipath propagation. MIMO is fundamentally different from smart antenna techniques developed to enhance the performance of a single data signal, such as beamforming and diversity.
If someone asks the question, "How communication happens between a transmitter and a receiver?", one will typically say something like below that there is a transmitter with antenna sending the signal. The signal transmits through the channel and there is a receiver that the same frequency, receives and then decodes the signal.
Suppose, consider a case where there are multiple antennas at the transmitter(say M) and multiple antennas at the receiver(say N). Each antenna at the transmitter is used to send different data blocks on the same spectrum. Will the receiver will be able to decode it? One could say that it may not. Since both antennas are using the same bandwidth spectrum, the interference can be high so that the receiver won't be able to decode the signals. But in MIMO, at the expense of additional hardware and computation complexity, we can increase the number of parallel data streams with the same available bandwidth. That means, with MIMO, we can actually increase the channel capacity.
A key feature of MIMO systems is the ability to turn multipath propagation, traditionally a pitfall of wireless transmission, into a benefit for the user. In such a MIMO system, there are M x N signal paths from the transmit antennas to the receive antennas, and the signals on these paths are not identical. The signal component at each antenna at the receiver includes the signal component from all transmit antennas. Since the MxN multi-paths are different, the channel fading seen at each of the receiver antenna too is different. The signal at the receiver can be unscrambled only when it has good transmit diversity (due to man-made objects or natural obstacles). This diversity in the arrived signal makes it possible for the DSP to unscramble the signal, decode the transmitted data blocks and then makes it into a parallel stream. In some cases, the receiver can send the channel estimation info to the transmitter to adjust the transmission. Such kind of system is called a closed-loop MIMO system.
We should then ask the question - for a MIMO system with M transmit antennas and N receive antennas, how many such parallel data streams are possible? the answer is clear - it is the minimum of (M, N). So, at maximum, under ideal conditions (with a rich multipath environment), it is possible to have min(M, N) times the channel capacity of the SISO system.
Is it always possible to have channel capacity multiplied by min(M, N)? The answer is NO. For MIMO to work, it should have multiple data streams that are uncorrelated in nature. In the case of the MxN system, if there exists LOS between transmitter and receiver, then the no of orthogonal data streams that can be transmitted will come down to 1 as the paths between tx and rx antennas are almost same i.e. the correlation is high. Basically, as the multipath between Tx & Rx decreases, the no of uncorrelated data streams also decreases. The beauty of MIMO is that we are using the naturally available multipath condition(which is considered as a pitfall) to our advantage.
I tried to avoid mathematics to explain the MIMO concept. If you really want to know how exactly it works with mathematics, Please refer to the below IEEE document & chapter 5 in the below book.
Reference: From theory to practice: an overview of MIMO space-time coded wireless systems
Book: https://books.google.co.in/books?id=G5C5ii8O_y0C&printsec=frontcover#v=onepage&q&f=false