NOMA is not used in 5G because it is an outdated technology, replaced with Massive MIMO.

The answers that you seek are found in Chapter 6 of my book "Introduction to Multiple Antenna Communications and Reconfigurable Surfaces":

https://www.nowpublishers.com/article/BookDetails/9781638283140

The short answer is that NOMA doesn't increase the system capacity, but it gives extra flexibility in how to divide the system capacity between the users.

Non-Orthogonal Multiple Access (NOMA) is a promising technology for 5G networks that enhances spectral efficiency and user connectivity by allowing multiple users to share the same frequency channel. This is achieved by superimposing the signals of multiple users at different power levels and decoding them using successive interference cancellation (SIC). The power allocation strategy in NOMA is crucial to its performance and involves several key principles:

Power Allocation Principles in NOMA

1. Channel State Information (CSI):

- Power allocation in NOMA relies heavily on the Channel State Information (CSI) of the users. Typically, users with better channel conditions (stronger CSI) are allocated less power, while users with poorer channel conditions (weaker CSI) are allocated more power. This ensures that all users can reliably decode their signals.

2. Successive Interference Cancellation (SIC):

- In SIC, the receiver decodes the signals in a sequential manner, starting with the strongest signal. The power allocation must be such that the signal intended for the user with the strongest channel is received with enough power to be decoded first and then subtracted from the combined signal. This process continues until all users' signals are decoded.

3. Fairness and Quality of Service (QoS):

- The power allocation strategy must balance the trade-off between system throughput and fairness among users. Ensuring a minimum Quality of Service (QoS) for all users is essential. Power allocation algorithms often incorporate fairness constraints to prevent any user from experiencing excessively low data rates.

Power Allocation Strategies

1. Fixed Power Allocation:

- A simple strategy where fixed power levels are assigned to users regardless of their channel conditions. While easy to implement, this strategy may not be optimal in terms of spectral efficiency and user fairness.

2. Dynamic Power Allocation:

- Power levels are dynamically adjusted based on real-time CSI and user requirements. This approach aims to optimize system performance by adapting to changing channel conditions and user demands.

3. Optimization-Based Approaches:

- Water-Filling Algorithm:

- A popular optimization technique where power is allocated based on the inverse of the channel gains, filling "water" into the "holes" of the users' channels. This method maximizes the sum rate but may need modifications for fairness.

- Convex Optimization:

- Formulating the power allocation problem as a convex optimization problem, where the objective is to maximize system throughput or minimize power consumption while satisfying user QoS constraints. Techniques such as Lagrange multipliers and dual decomposition are often used.

4. Game-Theoretic Approaches:

- Using game theory to model the power allocation problem as a game where users or base stations compete for resources. Equilibrium concepts like Nash Equilibrium can be applied to find stable and fair power allocation solutions.

Example: Downlink NOMA Power Allocation

In a downlink NOMA scenario with a base station serving two users (User A with strong CSI and User B with weak CSI), the power allocation could follow these steps:

1. CSI Feedback:

- Users provide their CSI to the base station.

2. Initial Power Allocation:

- Assign initial power levels based on CSI. For example, User A (strong CSI) gets less power, and User B (weak CSI) gets more power.

3. Superposition Coding:

- The base station superimposes the signals for User A and User B. The combined signal is transmitted over the shared channel.

4. SIC at User A:

- User A first decodes User B's signal (due to its higher power allocation), subtracts it from the combined signal, and then decodes its own signal.

5. Direct Decoding at User B:

- User B directly decodes its own signal, as it was allocated more power.

Practical Considerations

1. Interference Management:

- Effective interference management is essential to ensure that SIC works correctly. Power levels must be carefully chosen to maintain an acceptable Signal-to-Interference-plus-Noise Ratio (SINR) for all users.

2. User Grouping:

- Proper user grouping based on their CSI can enhance NOMA performance. Grouping users with similar channel conditions can simplify power allocation and improve overall efficiency.

3. Computational Complexity:

- While dynamic and optimization-based power allocation strategies offer better performance, they also introduce higher computational complexity. Efficient algorithms and real-time processing capabilities are necessary for practical implementation.

Power allocation in NOMA for 5G networks involves a careful balance of maximizing system throughput, ensuring user fairness, and maintaining QoS. By leveraging CSI, dynamic adjustments, and advanced optimization techniques, NOMA can significantly enhance the spectral efficiency and user connectivity in 5G networks. As 5G technology continues to evolve, refining these power allocation strategies will be crucial for unlocking the full potential of NOMA.