User Scheduling and Passive Beamforming for FDMA/OFDMA in Intelligent Reflection Surface

Wei Jiang,Hans D. Schotten

2023-03-16

Abstract:Most prior works on intelligent reflecting surface (IRS) merely consider point-to-point communications, including a single user, for ease of analysis. Nevertheless, a practical wireless system needs to accommodate multiple users simultaneously. Due to the lack of frequency-selective reflection, namely the set of phase shifts cannot be different across frequency subchannels, the integration of IRS imposes a fundamental challenge to frequency-multiplexing approaches such as frequency-division multiple access (FDMA) and the widely adopted technique called orthogonal FDMA (OFDMA). It motivates us to study (O)FDMA-based multi-user IRS communications to clarify which user scheduling and passive beamforming are favorable under this non-frequency-selective reflection environment. Theoretical analysis and numerical evaluation reveal that (O)FDMA does not need user scheduling when there are a few users. If the number of users becomes large, neither user scheduling nor IRS reflection optimization is necessary. These findings help substantially simplify the design of (O)FDMA-based IRS communications.

Information Theory,Signal Processing

What problem does this paper attempt to address?

### Problems Addressed by the Paper

The paper primarily investigates the issues of multi-user scheduling and passive beamforming in Intelligent Reflecting Surface (IRS)-assisted Frequency Division Multiple Access (FDMA) and Orthogonal Frequency Division Multiple Access (OFDMA) systems. Specifically, the paper aims to address the following key problems:

1. **Challenges of IRS Integration in Multi-User Communication**:
- Traditional IRS research mostly focuses on point-to-point communication scenarios, i.e., with only one user, to simplify the analysis. However, actual wireless systems need to support multiple users simultaneously.
- Since IRS cannot achieve frequency-selective reflection (i.e., the phase shift on different frequency sub-channels cannot be different), this poses fundamental challenges to frequency reuse technologies such as FDMA and OFDMA.

2. **Optimization of User Scheduling and Passive Beamforming**:
- In a non-frequency-selective reflection environment, what are the optimal user scheduling and passive beamforming strategies?
- How to design these strategies in a multi-user IRS system to maximize the system's spectral efficiency?

3. **Theoretical Analysis and Numerical Evaluation of System Performance**:
- Through theoretical analysis and numerical simulations, study the impact of different user scheduling and passive beamforming strategies on system performance.
- Pay special attention to the performance trends when the number of users is small and large.

### Main Findings

- **When the number of users is small**:
- The FDMA-IRS system does not require user scheduling. That is, when the number of users is small, there is no need to specifically select a user for optimization, and the system can still maintain high performance.

- **When the number of users is large**:
- Neither user scheduling nor IRS reflection optimization is needed. As the number of users increases, most users will be affected by phase-inconsistent reflections, so randomly selecting users or randomly setting phase shifts can also achieve near-optimal performance.

### Conclusion

The research results of the paper indicate that in IRS-assisted FDMA systems, when the number of users is small, the complex processes of user scheduling and reflection optimization can be omitted; and when the number of users is large, even simple random strategies can achieve near-optimal performance. These findings help significantly simplify the design of IRS-assisted FDMA systems, avoiding complex cascaded channel estimation and IRS tuning, thereby reducing system complexity and cost.