TL;DR
This paper tackles the challenge of integrating Device-to-Device (D2D) communication into cellular networks while simultaneously enabling Integrated Sensing and Communication (ISAC) capabilities at the base station (BS). They propose a novel joint transceiver beamforming and power allocation scheme for a full-duplex ISAC-enabled D2D underlaid cellular network, aiming to maximize the overall sum rate for both communication and sensing.
What Problem Does This Paper Address?
The increasing demand for high-quality communication and precise sensing in future wireless networks (like 6G) necessitates efficient spectrum utilization. Integrating D2D communication into cellular networks offers a way to reduce the burden on base stations and improve spectral efficiency. Simultaneously, ISAC systems aim to use the same hardware and spectrum for both communication and sensing. However, combining these two advancements – D2D underlaid cellular networks and ISAC – presents significant challenges. Specifically, managing interference between cellular users, D2D pairs, and the ISAC BS, while optimizing both communication and sensing performance, is a complex problem. Existing works often focus on either communication or sensing, or they don't consider the full-duplex nature of the ISAC BS or the complexities of a D2D underlaid scenario. This paper addresses the need for a unified approach that jointly optimizes beamforming and power allocation to maximize spectral efficiency in such a complex environment.
Methods
The core of this paper lies in its proposed optimization framework. They formulate a sum rate maximization problem for a full-duplex ISAC-enabled D2D underlaid cellular network. This problem is inherently non-convex due to the coupling of variables and the nature of the objective function and constraints. To tackle this, they employ a Successive Convex Approximation (SCA) based iterative algorithm. This method iteratively approximates the non-convex parts of the problem with convex functions, allowing them to solve a sequence of convex subproblems.
Key methodological contributions include:
- Joint Optimization: The scheme jointly optimizes the transmit beamforming vectors for the BS, the receive beamforming vector at the BS, and the transmit power of the D2D transmitters.
- Full-Duplex ISAC BS: The model considers a full-duplex BS, which can transmit and receive simultaneously, adding complexity due to self-interference.
- SCA Algorithm: The use of SCA is crucial for transforming the non-convex problem into a solvable form. They also provide a proof of convergence for their algorithm.
- Radar SINR Constraint: A critical aspect is the inclusion of a minimum sensing Signal-to-Interference-plus-Noise Ratio (SINR) constraint, ensuring adequate sensing performance.
Datasets
Not discussed in the paper. The evaluation is based on simulated scenarios with defined system parameters.
Key Insights and Ideas
- Synergy of ISAC and D2D: The paper demonstrates that effectively integrating ISAC capabilities into a D2D underlaid cellular network can lead to significant performance gains, outperforming schemes that focus solely on communication or sensing.
- Importance of Joint Optimization: The non-convexity of the problem highlights that optimizing communication and sensing independently would likely lead to suboptimal solutions. Jointly optimizing beamforming and power allocation is essential for achieving the best overall performance.
- SCA as a Powerful Tool: The successful application of the SCA algorithm to this complex, non-convex problem showcases its effectiveness in solving challenging wireless communication and sensing optimization tasks.
- Beampattern Control: The analysis of transmit and receive beampatterns reveals how the proposed scheme intelligently steers energy towards the target and communication users while simultaneously suppressing interference, a key advantage over simpler schemes.
Key Results
- Performance Improvement: The proposed scheme significantly outperforms benchmark schemes (Communication-only, Sensing-only, MRT, and ZF) in terms of both sum rate (communication performance) and radar SINR (sensing performance).
- Convergence: The SCA-based algorithm converges quickly, typically within 5 iterations, even with a large number of antennas (e.g., 32x32).
- Rank-One Solution Ratio: The iterative optimization process for the beamforming matrices (Wk) yields a high ratio of rank-one solutions (over 99.2%), which is desirable for practical implementation.
- Slight Performance Trade-off: Integrating radar functionality leads to a slight performance loss in communication compared to a communication-only scheme, but this loss is minimal and well-compensated by the dual functionality.
- Interference Suppression: The proposed scheme effectively suppresses interference from the BS to D2D users and also from D2D users to the BS, which is crucial for the stability of the system.
Limitations and Future Work
- Rank-One Constraint Relaxation: While the SCA algorithm is applied to a relaxed version of the problem (omitting the strict rank-one constraint on Wk), the results show a high ratio of rank-one solutions, suggesting the relaxation is effective. However, explicitly handling the rank-one constraint in a more direct manner could be a future direction.
- Channel State Information (CSI): The paper assumes perfect CSI. In real-world scenarios, imperfect CSI can degrade performance. Future work could explore robust optimization techniques under imperfect CSI.
- Hardware Impairments: The analysis does not explicitly account for hardware impairments like non-linear power amplifiers or phase noise, which are common in practical systems.
- Scalability: While the SCA algorithm shows good convergence, its computational complexity might still be a concern for extremely large-scale systems. Exploring more efficient distributed or machine learning-based approaches could be beneficial.
- Dynamic Environments: The current model assumes static users and targets. Extending the framework to dynamic environments with moving users and targets would be a valuable next step.
Summary updated