ICACT20250012 Slide.01
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
Hi everybody, my name is Nan-Hsiung Huang, and I am a PhD student.
Today¡¯s topic is Joint SIC and SCMA for NOMA-Based Uplink Transmission in 5G Systems
|
ICACT20250012 Slide.02
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
Here is the outline: Introduction, System Model, Review of Data Detection, Proposed Data Detection Schemes and Power Control Strategies, Simulations and Conclusions. |
ICACT20250012 Slide.03
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
5G has many scenarios. mMTC scenario needs more multi-access resources to allow many devices to connect.
NOMA technology meets this need. Both power-domain NOMA and SCMA offer higher access capacity.
PD-NOMA uses SIC at the receiver to detect users, while SCMA relies on MPA for multi-user detection.
|
ICACT20250012 Slide.04
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
To improve the traditional NOMA system, a hybrid approach is used, as well as the iterative detection and decoding (IDD) method for LDPC-coded SCMA systems to improve reliability.
However, most research focused on downlink multiple access transmission.
In this study, we focus on uplink multiple access and propose the SCMA-SIC process, which integrates IDD and SIC concepts. We also compare the performance of different NOMA receiver designs, including IDD, hybrid-domain NOMA, and our proposed schemes.
Finally, three different power control strategies are used to improve system performance.
|
ICACT20250012 Slide.05
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
In this study, we consider a multi-user uplink system with a base station (BS) located at the center. |
ICACT20250012 Slide.06
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
The picture shows the SCMA encoder with 6 users transmitting on 4 resource elements. |
ICACT20250012 Slide.07
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
The block diagram of the uplink coded SCMA transmitter
The received data signal y given by equation (1)
x_j is the SCMA codeword for user j, h_j is the channel vector, n is the white noise vector.
|
ICACT20250012 Slide.08
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
In receiver, it is assumed that the BS has accurate information of active codebooks (CBs) and channel gain. |
ICACT20250012 Slide.09
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
As shown in the figure, IDD uses an iterative method for data detection for coded SCMA.
where Nout is the number of outer iterations, Ns is the number of SCMA iterations, and NL is the number of LDPC iterations.
|
ICACT20250012 Slide.10
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
As shown in Fig, The hybrid-domain NOMA for data detection.
First, it decodes the stronger users and then performs SIC on them. Finally, it decodes the weaker users.
This method has low complexity, but the power gap between stronger and weaker users affects system performance.
|
ICACT20250012 Slide.11
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
This is the flowchart of the proposed SCMA-SIC process.
The MPA detector first detects the strongest user, which is then decoded by the LDPC decoder. Using the SIC technique, the detected user's signal is subtracted from the received signal, and the remaining users are sequentially decoded according to this process.
|
ICACT20250012 Slide.12
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
This is the flowchart of the proposed SCMA-SIC group process.
Compared to the SCMA-SIC process, this approach first requires grouping users based on their received signal power. Subsequent iterations are then performed on a group-by-group.
While this scheme reduces complexity compared to SCMA-SIC, it may result in some performance loss.
|
ICACT20250012 Slide.13
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
Strategy A: With power control for equal received power
Strategy B: Without power control
Strategy C: With power control and different levels of power adjustment
|
ICACT20250012 Slide.14
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
Strategy A achieves equal received power through power control.
P_(R,j): equal received power , P_Toatl: Total transmit power
Strategy B sets uniform transmit power across users (i.e., no power control)
|
ICACT20250012 Slide.15
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
Strategy C involves power control with different levels of power adjustment
The transmit power is given by P_(T,j)^C
P_(T,j)^C' from dB to the linear scale before normalizing it.
P_(T,j)^C' is set to be the sum of the baseline power P_(T,j)^A (in dB).
P_(L,j)^C ¡¿ Pstep is the power adjustment term.
P_step: the step size (in dB)
|
ICACT20250012 Slide.16
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
Key simulation parameters include uplink transmission, UE randomly chooses the CB, the average probability of connection and BLER performance. |
ICACT20250012 Slide.17
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
Compare the system performance with and without CB collision, taking 4 active UEs as an example.
Even with CB collision, decoding is still possible, though there is a slight drop in performance. So, all following simulations include CB collisions.
|
ICACT20250012 Slide.18
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
The fig. is BLER with 2000 UEs, P = 0.003, L = 2, and Pstep = 10 dB (without power control).
The proposed SCMA-SIC process achieves the best performance compared to others. In hybrid-domain NOMA, the performance of stronger users drops once Eb/N0 exceeds 10 dB due to the power gap between stronger and weaker users. Therefore, it will not be included in further simulations.
|
ICACT20250012 Slide.19
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
The fig. is BLER with 2000 UEs, P = 0.003, L = 3, and Pstep = 3 dB, and with different power control strategies.
The SCMA-SIC process consistently outperforms others, with strategy C showing the best performance.
|
ICACT20250012 Slide.20
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
We proposed the receiver for coded SCMA systems in uplink multiple access transmission, which improves the accuracy of multi-user information decoding.
The SCMA-SIC and SCMA-SIC-G processes are compared with existing data detection schemes in terms of BLER. The simulation results show that the SCMA-SIC process is best.
When power control strategy C is adopted, the SCMA-SIC process always achieves the best BLER.
|
ICACT20250012 Slide.21
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
Thanks for listening! I'm happy to share our research with you. |
ICACT20250012 Slide.22
[Big Slide]
[YouTube] |
Chrome Click!! |
 |
References |