Home >
Community >
In NOMA, more power allocation to the users with poor channels, is...
Upvote
28
Downvote
+ Scheme
Posted by
L.M. Lalko
In NOMA, more power allocation to the users with poor channels, is...
Totally agree with Prof. Mojtaba Vaezi that the sum rate and user fairness cannot be increased simultaneously and hence power should be allocated depending on the requirement.
Totally agree with Prof. Mojtaba Vaezi that the sum rate and user fairness cannot be increased simultaneously and hence power should be allocated depending on the requirement.
Let us consider a scenario where two users exist i.e. a near-end and far-end users. Therefore, the transmitter will transmit the information into two identical phases. Suppose, in the first transmission phase, it transmits the information to a near-end user, and then in the second phase, it transmits the information to the far-end user. In particular, channel conditions are different in each transmission phase. Therefore, I just want to know, how we can directly add two rates, which are obtained under different channel conditions, in order to find the sum rate for end-to-end transmission. As per my understanding, the suggested paper is not concerned with the different channel conditions.
Let us consider a scenario where two users exist i.e. a near-end and far-end users. Therefore, the transmitter will transmit the information into two identical phases. Suppose, in the first transmission phase, it transmits the information to a near-end user, and then in the second phase, it transmits the information to the far-end user. In particular, channel conditions are different in each transmission phase. Therefore, I just want to know, how we can directly add two rates, which are obtained under different channel conditions, in order to find the sum rate for end-to-end transmission. As per my understanding, the suggested paper is not concerned with the different channel conditions.
My question is not about symmetric & asymmetric channels, which is related to imperfect channel conditions. I just want to understand how the sum rate is obtained. It is defined as the direct addition of two different channel capacities (OR rates) which are obtained for two different users under different channel conditions.
My question is not about symmetric & asymmetric channels, which is related to imperfect channel conditions. I just want to understand how the sum rate is obtained. It is defined as the direct addition of two different channel capacities (OR rates) which are obtained for two different users under different channel conditions.
NOMA cannot increase the sum rate compared to OMA if you have no restriction on the number of users. Assume you have a bandwidth W and you want to increase the sum-rate but the number of users is not important. Consider the following two cases: 1. You allocate all bandwidth (W) to serve only one user (which results in OMA) 2. You allocate the bandwidth (W) to two users according to NOMA broadcasting
Then, the first case always has a better sum rate regardless of power allocation or channel conditions. Please read the paper carefully. This is elaborated on in Table 1.
However, if you for example divide W into two parts and allocate one user in each part (you are mandating to have two users) then the sum rate will be less than NOMA. So, simply saying NOMA has a better sum rate than OMA is wrong. It can have a better sum rate only if you want to serve the same number of users.
NOMA cannot increase the sum rate compared to OMA if you have no restriction on the number of users. Assume you have a bandwidth W and you want to increase the sum-rate but the number of users is not important. Consider the following two cases: 1. You allocate all bandwidth (W) to serve only one user (which results in OMA) 2. You allocate the bandwidth (W) to two users according to NOMA broadcasting
Then, the first case always has a better sum rate regardless of power allocation or channel conditions. Please read the paper carefully. This is elaborated on in Table 1.
However, if you for example divide W into two parts and allocate one user in each part (you are mandating to have two users) then the sum rate will be less than NOMA. So, simply saying NOMA has a better sum rate than OMA is wrong. It can have a better sum rate only if you want to serve the same number of users.
Let's say the far user has a weak channel and less power is allocated, according to the Shannon capacity formula, the achievable maximum data rate becomes
B*log_2(1+h_{weak}^2*P_{weak})>? R_{req} (1)
The problem is whether (1) satisfies your data rate requirement or not. If the data rate requirement is low, then the UE may send the same data repeatedly to make sure a reliable communication.
Also for further reading, you may have a look at this article [1].
[1] Z. Ding, R. Schober, and H. V. Poor “Unveiling the Importance of SIC in NOMA Systems - Part I - State of the Art and Recent Findings ”, IEEE Communication Letters
Let's say the far user has a weak channel and less power is allocated, according to the Shannon capacity formula, the achievable maximum data rate becomes
B*log_2(1+h_{weak}^2*P_{weak})>? R_{req} (1)
The problem is whether (1) satisfies your data rate requirement or not. If the data rate requirement is low, then the UE may send the same data repeatedly to make sure a reliable communication.
Also for further reading, you may have a look at this article [1].
[1] Z. Ding, R. Schober, and H. V. Poor “Unveiling the Importance of SIC in NOMA Systems - Part I - State of the Art and Recent Findings ”, IEEE Communication Letters
This is a great question, and as you correctly mentioned earlier this is not true and is debunked in Myth 1 in the following paper.
M. Vaezi, R. Schober, Z. Ding, and H. V. Poor, ‘‘Non-orthogonal multiple access: Common myths and critical questions,’’ IEEE Wireless Communications, vol. 26, no. 5, pp. 174–180, October 2019. http://www.ece.villanova.edu/~mvaezi/papers/noma_myths19.pdf
Then, to answer your question about "what about the published work?" I should say unfortunately some early highly-cited works of NOMA made this incorrect assumption and that spread to many other works. The goal of the above paper is to debunk such myths.
You may also find the following interesting:
The main purpose of NOMA in 5G and beyond is to increase the number of users. Besides that
1. If you want to increase the sum-rate of the network you should allocate more power to the users with 'strong' channels (quite the opposite of what is suggested in some papers, and stated in your question). Please refer to Table 4 and Myth 4 in the above paper.
2. If your goal is to improve user fairness then you should allocate more power to the users with 'poor' channels. However, you should note that with such power allocation you may largely reduce the sum rate and spectral efficiency compared to OMA. This paower allocation is as if you make the richs poor so that every one is poor, and call it fairness!
In short, you cannot increase the sum rate and user fairness, simultaneously. These are related reversely.
This is a great question, and as you correctly mentioned earlier this is not true and is debunked in Myth 1 in the following paper.
M. Vaezi, R. Schober, Z. Ding, and H. V. Poor, ‘‘Non-orthogonal multiple access: Common myths and critical questions,’’ IEEE Wireless Communications, vol. 26, no. 5, pp. 174–180, October 2019. http://www.ece.villanova.edu/~mvaezi/papers/noma_myths19.pdf
Then, to answer your question about "what about the published work?" I should say unfortunately some early highly-cited works of NOMA made this incorrect assumption and that spread to many other works. The goal of the above paper is to debunk such myths.
You may also find the following interesting:
The main purpose of NOMA in 5G and beyond is to increase the number of users. Besides that
1. If you want to increase the sum-rate of the network you should allocate more power to the users with 'strong' channels (quite the opposite of what is suggested in some papers, and stated in your question). Please refer to Table 4 and Myth 4 in the above paper.
2. If your goal is to improve user fairness then you should allocate more power to the users with 'poor' channels. However, you should note that with such power allocation you may largely reduce the sum rate and spectral efficiency compared to OMA. This paower allocation is as if you make the richs poor so that every one is poor, and call it fairness!
In short, you cannot increase the sum rate and user fairness, simultaneously. These are related reversely.
In terms of the answer on Feb. 4, here are some of my considerations. Adding two rates in your case means a sum rate. The achievable rate pair of the NOMA or Broadcast channel is r1 ≤ log(1 + αP|h1|^2/N0), r2 ≤ log(1 + (1 − α)P|h2|^2/( αP|h2| ^2 + N0), where h1 and h2 are the channels. See (5) in [1] for more details, and [2, Chapter 6.2.2, pp. 279].
When you mentioned the different channel conditions, do you mean the channels could have different relations and values? It may be the topic of symmetric and asymmetric channels. For example, in the suggested paper, Figure 2, the channel |h1| = 10 |h2|, the asymmetric channel is considered.
[1] Qi, Yue, Xinliang Zhang, and Mojtaba Vaezi. "Over-the-Air Implementation of NOMA: New Experiments and Future Directions." IEEE Access 9 (2021): 135828-135844. https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9552864 [2] D. Tse and P. Viswanath, Fundamentals of Wireless Communication. Cambridge, U.K.: Cambridge Univ. Press, 2005.
In terms of the answer on Feb. 4, here are some of my considerations. Adding two rates in your case means a sum rate. The achievable rate pair of the NOMA or Broadcast channel is r1 ≤ log(1 + αP|h1|^2/N0), r2 ≤ log(1 + (1 − α)P|h2|^2/( αP|h2| ^2 + N0), where h1 and h2 are the channels. See (5) in [1] for more details, and [2, Chapter 6.2.2, pp. 279].
When you mentioned the different channel conditions, do you mean the channels could have different relations and values? It may be the topic of symmetric and asymmetric channels. For example, in the suggested paper, Figure 2, the channel |h1| = 10 |h2|, the asymmetric channel is considered.
[1] Qi, Yue, Xinliang Zhang, and Mojtaba Vaezi. "Over-the-Air Implementation of NOMA: New Experiments and Future Directions." IEEE Access 9 (2021): 135828-135844. https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9552864 [2] D. Tse and P. Viswanath, Fundamentals of Wireless Communication. Cambridge, U.K.: Cambridge Univ. Press, 2005.
Thank you very much Prof. Majtaba Varzi and Mr. Anad Jee, for your answers to my question. Now it is clear to me that the NOMA can be used for two different purposes i.e. to increase the sum rate and user fairness. However, there is again a doubt/question that how sum-rate can be a direct addition of two rates, which are obtained under two different channel conditions?
Thank you very much Prof. Majtaba Varzi and Mr. Anad Jee, for your answers to my question. Now it is clear to me that the NOMA can be used for two different purposes i.e. to increase the sum rate and user fairness. However, there is again a doubt/question that how sum-rate can be a direct addition of two rates, which are obtained under two different channel conditions?
Totally agree with Prof. Mojtaba Vaezi that the sum rate and user fairness cannot be increased simultaneously and hence power should be allocated depending on the requirement.
Totally agree with Prof. Mojtaba Vaezi that the sum rate and user fairness cannot be increased simultaneously and hence power should be allocated depending on the requirement.
More
VOTE
Dear Dr. Vaezi,
Let us consider a scenario where two users exist i.e. a near-end and far-end users. Therefore, the transmitter will transmit the information into two identical phases. Suppose, in the first transmission phase, it transmits the information to a near-end user, and then in the second phase, it transmits the information to the far-end user. In particular, channel conditions are different in each transmission phase. Therefore, I just want to know, how we can directly add two rates, which are obtained under different channel conditions, in order to find the sum rate for end-to-end transmission. As per my understanding, the suggested paper is not concerned with the different channel conditions.
Dear Dr. Vaezi,
Let us consider a scenario where two users exist i.e. a near-end and far-end users. Therefore, the transmitter will transmit the information into two identical phases. Suppose, in the first transmission phase, it transmits the information to a near-end user, and then in the second phase, it transmits the information to the far-end user. In particular, channel conditions are different in each transmission phase. Therefore, I just want to know, how we can directly add two rates, which are obtained under different channel conditions, in order to find the sum rate for end-to-end transmission. As per my understanding, the suggested paper is not concerned with the different channel conditions.
More
VOTE
Dear Yue Qi,
Thanks for your answer!
My question is not about symmetric & asymmetric channels, which is related to imperfect channel conditions. I just want to understand how the sum rate is obtained. It is defined as the direct addition of two different channel capacities (OR rates) which are obtained for two different users under different channel conditions.
Dear Yue Qi,
Thanks for your answer!
My question is not about symmetric & asymmetric channels, which is related to imperfect channel conditions. I just want to understand how the sum rate is obtained. It is defined as the direct addition of two different channel capacities (OR rates) which are obtained for two different users under different channel conditions.
More
VOTE
Ambrish,
NOMA cannot increase the sum rate compared to OMA if you have no restriction on the number of users. Assume you have a bandwidth W and you want to increase the sum-rate but the number of users is not important. Consider the following two cases:
1. You allocate all bandwidth (W) to serve only one user (which results in OMA)
2. You allocate the bandwidth (W) to two users according to NOMA broadcasting
Then, the first case always has a better sum rate regardless of power allocation or channel conditions. Please read the paper carefully. This is elaborated on in Table 1.
However, if you for example divide W into two parts and allocate one user in each part (you are mandating to have two users) then the sum rate will be less than NOMA. So, simply saying NOMA has a better sum rate than OMA is wrong. It can have a better sum rate only if you want to serve the same number of users.
Ambrish,
NOMA cannot increase the sum rate compared to OMA if you have no restriction on the number of users. Assume you have a bandwidth W and you want to increase the sum-rate but the number of users is not important. Consider the following two cases:
1. You allocate all bandwidth (W) to serve only one user (which results in OMA)
2. You allocate the bandwidth (W) to two users according to NOMA broadcasting
Then, the first case always has a better sum rate regardless of power allocation or channel conditions. Please read the paper carefully. This is elaborated on in Table 1.
However, if you for example divide W into two parts and allocate one user in each part (you are mandating to have two users) then the sum rate will be less than NOMA. So, simply saying NOMA has a better sum rate than OMA is wrong. It can have a better sum rate only if you want to serve the same number of users.
More
VOTE
Dear all,
Let's say the far user has a weak channel and less power is allocated,
according to the Shannon capacity formula, the achievable maximum data rate becomes
B*log_2(1+h_{weak}^2*P_{weak})>? R_{req} (1)
The problem is whether (1) satisfies your data rate requirement or not.
If the data rate requirement is low, then the UE may send the same data repeatedly to make sure a reliable communication.
Also for further reading, you may have a look at this article [1].
[1] Z. Ding, R. Schober, and H. V. Poor “Unveiling the Importance of SIC in NOMA Systems - Part I - State of the Art and Recent Findings ”, IEEE Communication Letters
Best regards,
Dear all,
Let's say the far user has a weak channel and less power is allocated,
according to the Shannon capacity formula, the achievable maximum data rate becomes
B*log_2(1+h_{weak}^2*P_{weak})>? R_{req} (1)
The problem is whether (1) satisfies your data rate requirement or not.
If the data rate requirement is low, then the UE may send the same data repeatedly to make sure a reliable communication.
Also for further reading, you may have a look at this article [1].
[1] Z. Ding, R. Schober, and H. V. Poor “Unveiling the Importance of SIC in NOMA Systems - Part I - State of the Art and Recent Findings ”, IEEE Communication Letters
Best regards,
More
VOTE
Dear Ambrish,
This is a great question, and as you correctly mentioned earlier this is not true and is debunked in Myth 1 in the following paper.
M. Vaezi, R. Schober, Z. Ding, and H. V. Poor, ‘‘Non-orthogonal multiple access: Common myths and critical questions,’’ IEEE Wireless Communications, vol. 26, no. 5, pp. 174–180, October 2019. http://www.ece.villanova.edu/~mvaezi/papers/noma_myths19.pdf
Then, to answer your question about "what about the published work?" I should say unfortunately some early highly-cited works of NOMA made this incorrect assumption and that spread to many other works. The goal of the above paper is to debunk such myths.
You may also find the following interesting:
The main purpose of NOMA in 5G and beyond is to increase the number of users. Besides that
1. If you want to increase the sum-rate of the network you should allocate more power to the users with 'strong' channels (quite the opposite of what is suggested in some papers, and stated in your question). Please refer to Table 4 and Myth 4 in the above paper.
2. If your goal is to improve user fairness then you should allocate more power to the users with 'poor' channels. However, you should note that with such power allocation you may largely reduce the sum rate and spectral efficiency compared to OMA. This paower allocation is as if you make the richs poor so that every one is poor, and call it fairness!
In short, you cannot increase the sum rate and user fairness, simultaneously. These are related reversely.
Dear Ambrish,
This is a great question, and as you correctly mentioned earlier this is not true and is debunked in Myth 1 in the following paper.
M. Vaezi, R. Schober, Z. Ding, and H. V. Poor, ‘‘Non-orthogonal multiple access: Common myths and critical questions,’’ IEEE Wireless Communications, vol. 26, no. 5, pp. 174–180, October 2019. http://www.ece.villanova.edu/~mvaezi/papers/noma_myths19.pdf
Then, to answer your question about "what about the published work?" I should say unfortunately some early highly-cited works of NOMA made this incorrect assumption and that spread to many other works. The goal of the above paper is to debunk such myths.
You may also find the following interesting:
The main purpose of NOMA in 5G and beyond is to increase the number of users. Besides that
1. If you want to increase the sum-rate of the network you should allocate more power to the users with 'strong' channels (quite the opposite of what is suggested in some papers, and stated in your question). Please refer to Table 4 and Myth 4 in the above paper.
2. If your goal is to improve user fairness then you should allocate more power to the users with 'poor' channels. However, you should note that with such power allocation you may largely reduce the sum rate and spectral efficiency compared to OMA. This paower allocation is as if you make the richs poor so that every one is poor, and call it fairness!
In short, you cannot increase the sum rate and user fairness, simultaneously. These are related reversely.
More
VOTE
Thank Mr. Dusun for answer! I agree with you that reliability of the network at low rate can be enhanced by increasing the diversity order!
Thank Mr. Dusun for answer! I agree with you that reliability of the network at low rate can be enhanced by increasing the diversity order!
More
VOTE
Dear Colleagues,
I agree with you that the allocation of power depends on the requirement.
Dear Colleagues,
I agree with you that the allocation of power depends on the requirement.
More
VOTE
Hi, Ambrish,
In terms of the answer on Feb. 4, here are some of my considerations.
Adding two rates in your case means a sum rate.
The achievable rate pair of the NOMA or Broadcast channel is
r1 ≤ log(1 + αP|h1|^2/N0),
r2 ≤ log(1 + (1 − α)P|h2|^2/( αP|h2| ^2 + N0),
where h1 and h2 are the channels.
See (5) in [1] for more details, and [2, Chapter 6.2.2, pp. 279].
When you mentioned the different channel conditions, do you mean the channels could have different relations and values? It may be the topic of symmetric and asymmetric channels. For example, in the suggested paper, Figure 2, the channel |h1| = 10 |h2|, the asymmetric channel is considered.
[1] Qi, Yue, Xinliang Zhang, and Mojtaba Vaezi. "Over-the-Air Implementation of NOMA: New Experiments and Future Directions." IEEE Access 9 (2021): 135828-135844. https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9552864
[2] D. Tse and P. Viswanath, Fundamentals of Wireless Communication. Cambridge, U.K.: Cambridge Univ. Press, 2005.
I hope the discussion can make the topic clearer.
Best,
Yue
Hi, Ambrish,
In terms of the answer on Feb. 4, here are some of my considerations.
Adding two rates in your case means a sum rate.
The achievable rate pair of the NOMA or Broadcast channel is
r1 ≤ log(1 + αP|h1|^2/N0),
r2 ≤ log(1 + (1 − α)P|h2|^2/( αP|h2| ^2 + N0),
where h1 and h2 are the channels.
See (5) in [1] for more details, and [2, Chapter 6.2.2, pp. 279].
When you mentioned the different channel conditions, do you mean the channels could have different relations and values? It may be the topic of symmetric and asymmetric channels. For example, in the suggested paper, Figure 2, the channel |h1| = 10 |h2|, the asymmetric channel is considered.
[1] Qi, Yue, Xinliang Zhang, and Mojtaba Vaezi. "Over-the-Air Implementation of NOMA: New Experiments and Future Directions." IEEE Access 9 (2021): 135828-135844. https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9552864
[2] D. Tse and P. Viswanath, Fundamentals of Wireless Communication. Cambridge, U.K.: Cambridge Univ. Press, 2005.
I hope the discussion can make the topic clearer.
Best,
Yue
More
VOTE
Thank you very much Prof. Majtaba Varzi and Mr. Anad Jee, for your answers to my question. Now it is clear to me that the NOMA can be used for two different purposes i.e. to increase the sum rate and user fairness. However, there is again a doubt/question that how sum-rate can be a direct addition of two rates, which are obtained under two different channel conditions?
Thank you very much Prof. Majtaba Varzi and Mr. Anad Jee, for your answers to my question. Now it is clear to me that the NOMA can be used for two different purposes i.e. to increase the sum rate and user fairness. However, there is again a doubt/question that how sum-rate can be a direct addition of two rates, which are obtained under two different channel conditions?
More
VOTE