TL;DR
This paper presents a tutorial on Joint Communications and Sensing (JCAS) using OFDM signals, highlighting critical issues and proposing a novel, efficient sensing design. The core contribution is a decimation-based method that significantly reduces sensing complexity with minimal performance impact, making JCAS more practical for resource-constrained IoT systems.
What Problem Does This Paper Address?
The integration of communication and sensing functionalities into a single system, known as Joint Communications and Sensing (JCAS), holds immense promise for improving efficiency in Internet of Things (IoT) devices. Orthogonal Frequency Division Multiplexing (OFDM) is a popular waveform for both communication and sensing. However, existing OFDM sensing methods, particularly the Classical OFDM Sensing (COS) and its variant C-COS, face limitations. These include restricted sensing range due to cyclic prefix (CP) length, potential for noise enhancement, and a lack of comprehensive analysis for future OFDM variants like DFT-s-OFDM and OTFS. This paper aims to address these overlooked issues by providing a tutorial, reviewing recent progress, and introducing a novel design to enhance sensing efficiency.
Methods
The paper starts by reviewing existing OFDM sensing methods, primarily COS and C-COS, which process received signals in the frequency domain to generate a Range-Doppler Map (RDM). It then delves into recent research efforts addressing limitations like sensing range and noise enhancement. The core novel contribution is a Decimation-Based COS (DCOS) method. This approach leverages the observation that the processed echo signal in OFDM sensing often has a significantly smaller bandwidth than the original OFDM symbol. By applying efficient bandpass decimation (anti-aliasing filtering, downsampling, and frequency shifting), DCOS reduces the number of samples in the signal, leading to a smaller RDM. This, in turn, significantly reduces the computational complexity for subsequent target detection and estimation. The paper also discusses the application of these concepts to DFT-s-OFDM and OTFS waveforms.
Datasets
Not discussed in the paper. The paper relies on simulations using synthesized data based on defined signal models and parameters.
Key Insights and Ideas
- OFDM's Dual Role: OFDM, a ubiquitous communication waveform, is inherently suitable for sensing due to its structured nature, allowing for the extraction of target information from communication signals.
- Underlying Redundancy: The processed echo signal in OFDM sensing, particularly after point-wise division (PWD) or product (PWP), exhibits significant bandwidth redundancy. This redundancy is a key target for optimization.
- Decimation as a Complexity Reducer: Efficiently decimating the signal before RDM generation is a powerful technique to reduce computational load without substantial performance degradation.
- Trade-offs in PWD vs. PWP: While PWP (used in C-COS) can mitigate noise enhancement compared to PWD (used in COS), it can introduce a noise floor in high SNR scenarios. Analytical SINR comparisons are provided to guide selection.
- Future Waveform Compatibility: The proposed framework and decimation concept are designed to be adaptable to emerging OFDM variants like DFT-s-OFDM and OTFS, highlighting their forward-looking nature.
Key Results
- The proposed DCOS method significantly reduces the complexity of RDM generation by an order of magnitude compared to COS.
- Simulations demonstrate that DCOS achieves a comparable sensing performance (detection and estimation) to COS, with a slightly wider mainlobe in range but no impact on velocity measurement.
- The DCOS method shows a non-trivially lower wall-clock time for processing compared to C-COS, validating its reduced complexity.
- The paper highlights that conventional COS/C-COS methods fail to detect targets when the delay exceeds the CP length, a limitation that DCOS, by flexibly setting parameters, can overcome.
Limitations and Future Work
- Fake Targets in Novel Framework: The novel sensing framework proposed in Section 4 can generate "fake targets" due to the partial periodicity introduced by virtual CPs. Future work needs to address methods for differentiating fake from true targets or designing parameters to eliminate this issue.
- Scenario-Adaptive Parameter Selection: While core parameters of the sensing framework have been identified, a method for scenario-adaptive selection is still missing.
- MIMO JCAS: Extending current multi-/single-carrier JCAS methods to MIMO communications is challenging due to the non-orthogonal nature of signals from multiple antennas. Further investigation is needed.
- Hardware Imperfections: The impact of practical communication system components like pulse-shaping filters and other hardware imperfections on sensing performance is not fully accounted for and requires further study, potentially through prototype-assisted experiments.
- Analytical Model for C-COS CCC: An analytical model for the Cyclic Cross-Correlation (CCC) result in C-COS is lacking, making its sidelobe behavior unpredictable.
Summary updated