TL;DR
This paper proposes a novel approach for Integrated Sensing and Communication (ISAC) systems that leverages multipath components (MPCs) – typically a nuisance in wireless signals – to enhance both sensing accuracy and communication reliability. By developing a symbol-level fusion scheme (SFMC) and employing Khatri-Rao space-time (KRST) coding, the authors demonstrate significantly improved localization performance in challenging multipath environments.
What Problem Does This Paper Address?
Next-generation wireless systems, particularly those envisioned for 6G, aim to integrate sensing and communication capabilities (ISAC). A major hurdle in achieving high-accuracy sensing, especially in dense urban areas, is the severe interference caused by multipath components (MPCs). Traditionally, research has focused on mitigating or eliminating these MPCs. This paper challenges that paradigm by proposing to harness MPCs as a valuable resource. The core problem is how to effectively utilize the information embedded within these multipath signals to simultaneously improve the precision of sensing (like target localization) and the robustness of communication, particularly in environments where line-of-sight (LoS) paths are obstructed or weak.
Methods
The paper introduces a two-stage Multipath Component-Aided Signal Processing (SFMC) scheme:
- MPCs Separation Stage: This stage focuses on extracting the relevant sensing information from the received ISAC signals. After canceling the communication signal component, the remaining sensing information is modeled as a third-order tensor. A regularized alternating least squares (RALS) method is employed for tensor decomposition to separate the contributions of different multipath components. This is an advancement over standard alternating least squares (ALS) methods, aiming for faster convergence and better precision. The output of this stage provides estimates of angle and delay information for each multipath component.
- Symbol-Level Fusion Stage: This stage leverages the separated multipath information for localization. Virtual anchors (VAs) are constructed based on prior environmental knowledge and the estimated multipath parameters. Crucially, instead of fusing raw data (as in some prior methods), this scheme performs a symbol-level fusion of the angle and delay information. This involves creating angle and delay matching matrices, which are then fused to generate angle and delay fusion vectors. The peak of the combined fusion vector indicates the estimated target location.
To enhance communication and sensing performance in multipath environments, the paper also incorporates Khatri-Rao space-time (KRST) coding. This coding scheme is known for its ability to combat multipath fading and enable blind channel state information recovery, contributing to both reliable communication and improved sensing.
Datasets
Not discussed in the paper. The evaluation is performed through simulations using parameters provided in Table I.
Key Insights and Ideas
- Turning a Nuisance into a Feature: The most significant insight is the shift from treating multipath components as interference to utilizing them as a valuable source of information for enhanced sensing and communication.
- Symbol-Level Fusion Superiority: The proposed symbol-level fusion approach is shown to be more effective than existing data-level fusion methods, likely because it better exploits the structured information within the multipath components.
- RALS for Robust Tensor Decomposition: Employing the RALS method for tensor decomposition offers improved convergence and accuracy in separating multipath components compared to traditional ALS.
- KRST Coding Synergy: The integration of KRST coding complements the SFMC scheme by providing robust communication and sensing capabilities in challenging multipath conditions.
Key Results
- Localization Accuracy: The proposed SFMC scheme demonstrates significantly improved localization accuracy. With four multipath components, the 90% accuracy is 0.06 m, a substantial improvement over the existing scheme's 0.2 m.
- Robustness to MPCs: As the number of multipath components increases, the localization accuracy and anti-noise performance of the SFMC scheme improve.
- SNR Improvement: Compared to a traditional LoS-only localization scheme, the SFMC scheme requires approximately 4 dB less SNR to achieve the same localization error of 0.156 m.
- Communication Performance: Simulation results for Bit Error Rate (BER) indicate that while increasing MPCs or modulation order can raise BER (reducing reliability), increasing code length helps lower BER and improve communication performance.
Limitations and Future Work
- Stationary ISAC Systems: The current algorithm is primarily applicable to stationary ISAC systems. For mobile ISAC systems, an environmental map would need to be constructed beforehand to maintain the geometric relationship between the ISAC system and reflectors.
- Temporary MPCs: The paper acknowledges that the potential of temporary MPCs remains to be further explored, suggesting this as a direction for future work.
- Mobile ISAC Systems: Extending the SFMC scheme to handle mobile ISAC scenarios is a clear avenue for future research.
Summary updated