Spatial Modulation for Joint Radar-Communications Systems: Design, Analysis, and Hardware Prototype

Dingyou Ma, Nir Shlezinger, Tianyao Huang, Yariv Shavit, Moshe Namer, Yimin Liu, Yonina C. EldarIEEE Trans. Vehicular Technology 2021 · DOI

TL;DR

This paper introduces SpaCoR, a novel dual-function radar-communications (DFRC) system that cleverly uses spatial modulation (GSM) to allow a single antenna array to perform both radar and communication tasks simultaneously. By dynamically allocating antenna elements based on the transmitted communication data, SpaCoR achieves improved radar angular resolution and communication data rates compared to traditional fixed-allocation DFRC systems, as demonstrated by a functional hardware prototype.

What Problem Does This Paper Address?

Modern systems, like autonomous vehicles, increasingly require both radar for sensing and communication for data exchange. Traditionally, these functions are handled by separate, dedicated hardware, leading to increased size, cost, and power consumption. Dual-function radar-communications (DFRC) systems aim to consolidate these capabilities onto a single platform. However, existing DFRC approaches often face a trade-off: using communication waveforms for radar can degrade radar performance (e.g., high peak-to-average power ratio), while using radar waveforms for communication results in low data rates. Furthermore, joint optimization of waveforms can be computationally complex and require prior knowledge of communication channels and radar targets, which is often unavailable in dynamic environments. This paper addresses the challenge of achieving high performance in both radar and communication functionalities simultaneously, using a single antenna array, without requiring complex waveform design or extensive channel knowledge.

Methods

The core innovation of this paper is the Spatial Modulation based Communication-Radar (SpaCoR) system. SpaCoR leverages Generalized Spatial Modulation (GSM), a technique that embeds communication data not only in the transmitted signal's constellation but also in the selection of transmit antennas.

Here's how it works:

  • Dynamic Antenna Allocation: Instead of fixing a subset of antennas for radar and another for communication, SpaCoR dynamically assigns antenna elements to either function on a symbol-by-symbol basis. This allocation is determined by a portion of the communication data bits.
  • Joint Waveform: A single, combined Joint Radar-Communications (JRC) waveform is generated. This waveform carries both the radar signal and the communication symbols, with the communication symbols being transmitted from the dynamically selected antennas.
  • GSM for Communication: The communication subsystem uses GSM, which allows for higher data rates by encoding extra bits in the antenna selection. This means that for a given symbol duration, more information can be transmitted compared to traditional methods that only use constellation bits.
  • Radar Performance Enhancement: By using the full antenna array probabilistically over time (due to the random antenna selection for communication), the average transmit beam pattern of the radar subsystem closely approximates that of using the entire array solely for radar. This dynamic spatial agility improves angular resolution and reduces sidelobe levels.
  • Hardware Prototype: The authors built and demonstrated a hardware prototype of the SpaCoR system, validating its practical feasibility and performance.

Datasets

Not discussed in the paper. The evaluation relies on simulated data and experiments using a custom hardware prototype with controllable target scenarios.

Key Insights and Ideas

  • Spatial Modulation as a Unifying Principle: The paper elegantly demonstrates how spatial modulation, typically used for communication enhancement, can also be a powerful tool for improving radar performance by introducing spatial agility.
  • Trade-off Reversal: Instead of a trade-off between radar performance and communication data rate, SpaCoR leverages the communication data itself to improve radar performance (resolution, sidelobes) while simultaneously boosting communication rates.
  • Hardware Feasibility: The successful implementation of a hardware prototype validates the theoretical concepts and shows that SpaCoR is not just a theoretical construct but a practical solution.
  • Dynamic Allocation is Key: The dynamic, data-dependent allocation of antennas is the critical element that allows SpaCoR to achieve its dual benefits, unlike fixed allocation schemes.

Key Results

  • Communication Performance: SpaCoR achieved improved Bit Error Rate (BER) performance compared to fixed allocation DFRC schemes operating at the same data rate. This is attributed to GSM's ability to use sparser constellations by embedding bits in antenna selection.
  • Radar Angular Resolution: Experiments showed that SpaCoR significantly improved angular resolution, enabling the system to distinguish between closely spaced targets more effectively than fixed allocation methods (Fix1 and Fix2).
  • Sidelobe Reduction: The spatial agility induced by SpaCoR's transmission reduced sidelobe levels in the transmit beam pattern compared to fixed antenna allocations, leading to better performance in cluttered environments. The paper shows that SpaCoR's expected beam pattern approaches that of the full antenna array, outperforming Fix1 and Fix2.
  • Hardware Validation: The prototype successfully demonstrated over-the-air signaling, radar echo generation, and communication signal reception, confirming the practical viability of the proposed scheme.

Limitations and Future Work

  • Decoding Complexity: While GSM offers performance gains, it can increase decoding complexity at the receiver, especially for larger numbers of antennas or higher-order constellations. The paper mentions that reduced complexity GSM decoding methods exist and were considered for scenarios with larger R.
  • Dynamic Nature: The random nature of antenna allocation in SpaCoR, while beneficial for average performance, means that the instantaneous beam pattern can fluctuate. The paper analyzes the expected beam pattern and its variance.
  • Future Work: The paper suggests that further research could explore SpaCoR with sub-Nyquist sampling for radar, and investigate its performance in more complex scenarios with multiple targets and interference. The authors also highlight that their prototype can realize a multitude of DFRC systems, implying potential for exploring other DFRC strategies.

Summary updated

BibTeX
@article{ma2021spatial,
  author = {Dingyou Ma and Nir Shlezinger and Tianyao Huang and Yariv Shavit and Moshe Namer and Yimin Liu and Yonina C. Eldar},
  title = {{Spatial Modulation for Joint Radar-Communications Systems: Design, Analysis, and Hardware Prototype}},
  journal = {IEEE Trans. Vehicular Technology},
  year = {2021},
  doi = {10.1109/tvt.2021.3056408},
}
Figure