On the Performance of Non-Orthogonal Multiple Access in 5G Systems with Randomly Deployed Users

Abstract: In this letter, the performance of non-orthogonal multiple access (NOMA) is investigated in a cellular downlink scenario with randomly deployed users. The developed analytical results show that NOMA can achieve superior performance in terms of ergodic sum rates; however, the outage performance of NOMA depends critically on the choices of the users' targeted data rates and allocated power. In particular, a wrong choice of the targeted data rates and allocated power can lead to a situation in which the user's outage probability is always one, i.e. the user's targeted quality of service will never be met.

• The paper demonstrates that NOMA significantly improves spectral efficiency by leveraging power allocation and successive interference cancellation in 5G downlink scenarios.
• The paper reveals that proper power allocation under QoS constraints can effectively manage outage probability for users with fixed targeted rates.
• The paper derives ergodic sum rate approximations showing that, as user count increases, NOMA approaches the performance of opportunistic scheduling with enhanced fairness.

On the Performance of Non-Orthogonal Multiple Access in 5G Systems with Randomly Deployed Users

The paper "On the Performance of Non-Orthogonal Multiple Access in 5G Systems with Randomly Deployed Users" by Zhiguo Ding, Zheng Yang, Pingzhi Fan, and H. Vincent Poor investigates the performance of Non-Orthogonal Multiple Access (NOMA) under cellular downlink scenarios with randomly distributed users. This study elaborates on the benefits of NOMA in enhancing spectral efficiency, focusing on critical metrics such as ergodic sum rates and outage probability.

Introduction

The study begins by recognizing the transition from 4G mobile networks like LTE to emerging 5G networks, where NOMA has been identified as a promising technique due to its superior spectral efficiency. The authors examine NOMA performance in two scenarios: one where each user has a preset targeted data rate based on Quality of Service (QoS) requirements, and another where user rates are opportunistically allocated according to channel conditions.

NOMA Transmission Protocol

The transmission scenario considers a base station at the center of a disc of radius $\mathcal{R}_D$ with $M$ uniformly distributed users. The channel gain for each user, affected by Rayleigh fading and path loss, is represented as:

$h_m = \frac{\tilde{g}_m}{\sqrt{1+d_m^{\alpha}}}$

The base station transmits $\sum_{m=1}^{M} \sqrt{a_m P} s_m$, where $s_m$ is the message for the $m$-th user, $P$ is the transmission power, and $a_m$ is the power allocation coefficient. Users perform Successive Interference Cancellation (SIC) to decode messages. This allows for the derivation of the data rate achievable to each user, which depends on the targeted data rates $\tilde{R_j}$ and the power allocation $a_m$.

Case I: Outage Performance of NOMA

For the scenario where $\tilde{R}_m$ is fixed by QoS requirements, the outage probability $\mathrm{P}^{out}_{m}$ is studied. The analysis reveals that NOMA's performance is highly sensitive to the users' targeted rates and power allocation. A critical condition $a_j > \phi_j \sum_{i=j+1}^{M} a_i$ ensures that the outage probability does not become one. When this condition is satisfied, the diversity order achieved by NOMA for the $m$-th user is $\frac{1}{\rho^m}$, where $\rho$ is the SNR, demonstrating significant potential in terms of outage performance over conventional orthogonal methods.

Case II: Ergodic Sum Rate of NOMA

In the opportunistic rate allocation scenario, the authors derive the ergodic sum rate, providing an approximation for high SNR and studying the asymptotic behavior as the number of users grows indefinitely. They demonstrate that the sum rate achieved by NOMA is:

$$R_{sum} \approx \sum_{m=1}^{M-1} \log(1 + \frac{\rho |h_m|^2 a_m}{\rho |h_m|^2 \tilde{a}_m + 1}) + \log(1 + \rho |h_M|^2 a_M)$$

The analysis depicts that as $M \to \infty$, the ergodic sum rate approaches $\log(\rho \log \log M)$, matching the upper bound performance of an opportunistic user scheduling scheme while ensuring better user fairness.

Numerical Results

Simulation results corroborate the theoretical derivations, showing that NOMA outperforms random user scheduling orthogonal schemes in terms of both outage probability and ergodic sum rates. When configured correctly, NOMA ensures that all users experience the same diversity order as orthogonal access techniques.

Conclusions

The paper concludes that NOMA can provide significant enhancements in outage performance and ergodic sum rates compared to traditional orthogonal MA techniques. However, it also acknowledges the increased complexity due to SIC and lower gains at low SNRs, emphasizing the need for future research in optimizing performance-complexity trade-offs at varying SNR levels.

Implications and Future Directions

The implications of this research are twofold. Practically, it illustrates how NOMA can be implemented in 5G networks to improve spectral efficiency and user fairness. Theoretically, it sets a foundation for further exploration into optimal power allocation strategies and hybrid access schemes that combine NOMA with other promising techniques to optimize 5G network performance. Future work may also explore intelligent resource management levers facilitated by AI to dynamically adapt system parameters to user demographics and mobility patterns.