Abstract:In 5G networks, non-orthogonal multiple access (NOMA) provides a number of benefits by providing uneven power distribution to multiple users at once. On the other hand, effective power allocation, successful successive interference cancellation (SIC), and user fairness all depend on precise channel state information (CSI). Because of dynamic channels, imperfect models, and feedback overhead, CSI prediction in NOMA is difficult. Our aim is to propose a CSI prediction technique based on an ML model that accounts for partially decoded data (PDD), a byproduct of the SIC process. Our proposed technique has been shown to be efficient in handover failure (HOF) prediction and reducing pilot overhead, which is particularly important in 5G. We have shown how machine learning (ML) models may be used to forecast CSI in NOMA handover.
What problem does this paper attempt to address?
This paper attempts to solve the handover failure (HOF) and radio link failure (RLF) problems caused by inaccurate channel state information (CSI) prediction in non - orthogonal multiple access (NOMA) networks. Specifically:
1. **Problem Background**:
- In 5G networks, NOMA simultaneously serves multiple users by providing non - uniform power allocation, which brings many advantages.
- However, effective power allocation, successful successive interference cancellation (SIC), and user fairness all depend on accurate CSI.
- Due to dynamic channels, imperfect models, and feedback overhead, CSI prediction becomes very difficult in NOMA.
2. **Research Motivation**:
- Traditional CSI parameters such as RSRP, RSRQ, SNR, CQI, etc., although providing valuable channel information, are often unable to predict the dynamic behavior of wireless channels.
- Partially - decoded data (PDD), as a new CSI metric, can more realistically reflect channel performance and adapt to changes in user traffic.
3. **Solution**:
- A CSI prediction technique based on a machine - learning (ML) model is proposed, which utilizes partially - decoded data (PDD), a by - product of the SIC process.
- By analyzing the residual interference in PDD, future channel behavior can be better predicted, thereby improving SIC performance and reducing the probability of handover failure.
4. **Main Contributions**:
- Developed an ML model that uses PDD as an additional CSI metric to predict the channel state.
- Demonstrated the application of this CSI prediction model in reducing the handover failure rate, ping - pong effect, and the number of false positives.
- Demonstrated the superiority of this model in reducing pilot overhead through system - level simulation.
- Created artificial data to make up for the deficiency of the actual data set and demonstrated the competitive performance of the proposed model compared to other ML models.
5. **Key Formulas**:
- Received Signal Model:
\[
y=\sum_{i = 1}^{N}h_{i}x_{i}+n
\]
where \(y\) is the received signal, \(h_{i}\) is the channel matrix of the \(i\)-th user, \(x_{i}\) is the transmitted signal, and \(n\) is the additive white Gaussian noise.
- Normalized Root Mean Square Error (NRMSE):
\[
\text{NRMSE}=\frac{1}{\sigma_{\text{actual}}}\sqrt{\frac{1}{N}\sum_{i = 1}^{N}(y_{\text{predicted},i}-y_{\text{actual},i})^{2}}
\]
- Mean Absolute Scaled Error (MASE):
\[
\text{MASE}=\frac{\frac{1}{N}\sum_{i = 1}^{N}\vert y_{\text{predicted},i}-y_{\text{actual},i}\vert}{\frac{1}{N - 1}\sum_{i = 2}^{N}\vert y_{\text{actual},i}-y_{\text{actual},i - 1}\vert}
\]
- Mean Square Error (MSE):
\[
\text{MSE}=\frac{1}{N}\sum_{i = 1}^{N}(y_{\text{predicted},i}-y_{\text{actual},i})^{2}
\]
In conclusion, this paper aims to solve the problem of inaccurate CSI prediction in NOMA networks by introducing PDD as part of CSI and combining it with a machine - learning model, thereby improving network performance and user experience.