TL;DR
This paper introduces SIMRP, a novel method for mitigating self-interference in in-band full-duplex (IBFD) systems by jointly optimizing a Reconfigurable Intelligent Surface (RIS) and a receiver-side Phase Shifter Network (PSN). This approach effectively suppresses self-interference before it reaches the analog-to-digital converters (ADCs), significantly improving communication system performance and enabling higher data rates.
What Problem Does This Paper Address?
In-band full-duplex (IBFD) systems promise to double communication capacity by allowing simultaneous transmission and reception on the same frequency band. However, a major hurdle is the strong self-interference (SI) generated by the transmitter, which can overwhelm the receiver's analog-to-digital converters (ADCs). If this SI isn't mitigated before the ADCs, it saturates them, rendering subsequent digital signal processing useless and negating the benefits of IBFD. Existing methods have limitations, such as not being suitable for MIMO systems or reducing overall system capacity. This paper aims to address this by proposing a new technique that effectively mitigates SI without compromising system performance.
Methods
The core of the proposed SIMRP method lies in jointly optimizing two key components: a Reconfigurable Intelligent Surface (RIS) and a receiver-side Phase Shifter Network (PSN). The RIS, with its adjustable reflecting elements, and the PSN, deployed between the receiver antennas and the ADCs, work in tandem. The paper formulates the problem as minimizing the effective self-interference channel. Due to the non-convexity of the optimization problem, an alternating optimization approach is employed. This involves iteratively optimizing the RIS phases and the PSN coefficients. Specifically, the RIS phase optimization uses a Riemannian conjugate gradient (RCG) algorithm, while the PSN optimization utilizes an iterative sequential quadratic programming (SQP) method combined with a trust-region approach.
Datasets
Not discussed in the paper. The paper relies on simulated channel models for evaluation.
Key Insights and Ideas
- Synergistic RIS and PSN: The paper highlights that combining RIS and a PSN offers a powerful and flexible approach to self-interference mitigation, outperforming methods that rely on only one of these components.
- Pre-ADC Mitigation is Crucial: The work strongly reinforces the necessity of mitigating self-interference before the ADCs to prevent saturation and preserve signal integrity.
- Joint Optimization Unlocks Performance: The key insight is that jointly optimizing the phases of both the RIS and the PSN, rather than optimizing them independently, leads to significantly better self-interference suppression and, consequently, higher system sum rates.
- Maintaining Downlink Capacity: Unlike some prior RIS-assisted IBFD approaches that might reduce downlink capacity, SIMRP aims to preserve it by focusing solely on mitigating self-interference at the receiver.
Key Results
The simulation results demonstrate the effectiveness of the proposed SIMRP method:
- Significant Sum Rate Gain: SIMRP achieves a 33% gain in sum rate compared to a prior RIS-assisted IBFD (RAIBFD) system and a 70% gain over a conventional RIS-assisted FDD (RAFDD) system.
- Near-Ideal Performance: The proposed method reaches approximately 95% of the sum rate of an ideal IBFD system with infinite-bit ADCs.
- Robustness to ADC Resolution: Even with finite-bit ADCs (e.g., 8-bit), SIMRP maintains a substantial performance advantage.
- Robustness to Channel Estimation Error: The SIMRP method shows robustness to channel estimation errors, especially with larger RIS configurations and reasonably high ADC bit resolutions.
Limitations and Future Work
- Channel State Information (CSI): The proposed method assumes perfect CSI is known. The paper does investigate the impact of imperfect CSI in simulations, showing robustness, but a more in-depth analysis or adaptive CSI acquisition scheme could be a future direction.
- RIS and PSN Resolution: The paper considers RIS with finite bit resolutions and discusses its impact. Future work could explore more advanced RIS phase control mechanisms or adaptive bit allocation for both RIS and PSN.
- Hardware Complexity: While the paper focuses on performance, the practical implementation complexity and cost of integrating RIS and PSN in real-world systems could be a subject for future research.
- Scalability: The paper demonstrates performance with 8x8 RIS configurations. Investigating the scalability to even larger RIS arrays and more complex MIMO configurations would be valuable.
Summary updated