Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Over-the-Air Time-Frequency Synchronization in Distributed ISAC Systems

Kawon Han, Kaitao Meng, Christos Masouros

TL;DR

This work addresses the critical challenge of time-frequency synchronization in distributed ISAC systems by proposing an over-the-air framework that exploits offset reciprocity in reciprocal bistatic sensing channels. It introduces bistatic signal matching to compress multi-target information into a single offset and develops off-grid super-resolution estimators based on maximum likelihood and matrix pencil methods, with extensions to N-node centralized pairwise synchronization. Theoretical performance bounds via CRBs are provided, and a CPS strategy is developed to scale synchronization to large networks, with comprehensive simulations showing near-optimal target localization performance akin to fully synchronous systems. The results highlight the practical impact of precise synchronization on cooperative sensing and communication in D-ISAC, particularly in challenging NLoS environments. The framework paves the way for robust, scalable synchronization in future wireless networks that integrate distributed radar and communication functionalities.

Abstract

A distributed integrated sensing and communication (D-ISAC) system offers significant cooperative gains for both sensing and communication performance. These gains, however, can only be fully realized when the distributed nodes are perfectly synchronized, which is a challenge that remains largely unaddressed in current ISAC research. In this paper, we propose an over-the-air time-frequency synchronization framework for the D-ISAC system, leveraging the reciprocity of bistatic sensing channels. This approach overcomes the impractical dependency of traditional methods on a direct line-of-sight (LoS) link, enabling the estimation of time offset (TO) and carrier frequency offset (CFO) between two ISAC nodes even in non-LoS (NLOS) scenarios. To achieve this, we introduce a bistatic signal matching (BSM) technique with delay-Doppler decoupling, which exploits offset reciprocity (OR) in bistatic observations. This method compresses multiple sensing links into a single offset for estimation. We further present off-grid super-resolution estimators for TO and CFO, including the maximum likelihood estimator (MLE) and the matrix pencil (MP) method, combined with BSM processing. These estimators provide accurate offset estimation compared to spectral cross-correlation techniques. Also, we extend the pairwise synchronization leveraging OR between two nodes to the synchronization of $N$ multiple distributed nodes, referred to as centralized pairwise synchronization. We analyze the Cramer-Rao bounds (CRBs) for TO and CFO estimates and evaluate the impact of D-ISAC synchronization on the bottom-line target localization performance. Simulation results validate the effectiveness of the proposed algorithm, confirm the theoretical analysis, and demonstrate that the proposed synchronization approach can recover up to 96% of the bottom-line target localization performance of the fully-synchronous D-ISAC.

Over-the-Air Time-Frequency Synchronization in Distributed ISAC Systems

TL;DR

This work addresses the critical challenge of time-frequency synchronization in distributed ISAC systems by proposing an over-the-air framework that exploits offset reciprocity in reciprocal bistatic sensing channels. It introduces bistatic signal matching to compress multi-target information into a single offset and develops off-grid super-resolution estimators based on maximum likelihood and matrix pencil methods, with extensions to N-node centralized pairwise synchronization. Theoretical performance bounds via CRBs are provided, and a CPS strategy is developed to scale synchronization to large networks, with comprehensive simulations showing near-optimal target localization performance akin to fully synchronous systems. The results highlight the practical impact of precise synchronization on cooperative sensing and communication in D-ISAC, particularly in challenging NLoS environments. The framework paves the way for robust, scalable synchronization in future wireless networks that integrate distributed radar and communication functionalities.

Abstract

A distributed integrated sensing and communication (D-ISAC) system offers significant cooperative gains for both sensing and communication performance. These gains, however, can only be fully realized when the distributed nodes are perfectly synchronized, which is a challenge that remains largely unaddressed in current ISAC research. In this paper, we propose an over-the-air time-frequency synchronization framework for the D-ISAC system, leveraging the reciprocity of bistatic sensing channels. This approach overcomes the impractical dependency of traditional methods on a direct line-of-sight (LoS) link, enabling the estimation of time offset (TO) and carrier frequency offset (CFO) between two ISAC nodes even in non-LoS (NLOS) scenarios. To achieve this, we introduce a bistatic signal matching (BSM) technique with delay-Doppler decoupling, which exploits offset reciprocity (OR) in bistatic observations. This method compresses multiple sensing links into a single offset for estimation. We further present off-grid super-resolution estimators for TO and CFO, including the maximum likelihood estimator (MLE) and the matrix pencil (MP) method, combined with BSM processing. These estimators provide accurate offset estimation compared to spectral cross-correlation techniques. Also, we extend the pairwise synchronization leveraging OR between two nodes to the synchronization of multiple distributed nodes, referred to as centralized pairwise synchronization. We analyze the Cramer-Rao bounds (CRBs) for TO and CFO estimates and evaluate the impact of D-ISAC synchronization on the bottom-line target localization performance. Simulation results validate the effectiveness of the proposed algorithm, confirm the theoretical analysis, and demonstrate that the proposed synchronization approach can recover up to 96% of the bottom-line target localization performance of the fully-synchronous D-ISAC.

Paper Structure

This paper contains 25 sections, 5 theorems, 48 equations, 10 figures, 2 algorithms.

Table of Contents

  1. Introduction
  2. State-of-the-Art of Over-the-Air Synchronization
  3. Our Contributions
  4. System Model
  5. Signal Model for Over-the-Air Synchronization
  6. Sensing Channel Demodulation
  7. Pairwise Synchronization of Distributed ISAC Using Reciprocal Bistatic Sensing Channels
  8. Offset Reciprocity in Bistatic Observations
  9. Pre-processing for Delay-Doppler Decoupling
  10. Bistatic Signal Matching
  11. Super-resolution Offset Estimation for Distributed ISAC
  12. Maximum Likelihood Estimation
  13. Matrix Pencil
  14. Cramer-Rao Bound for Offset Estimation
  15. Centered Pairwise Synchronization of Distributed ISAC
  16. ...and 10 more sections

Key Result

Proposition 1

The TO and CFO between node $n$ and node $m$ in the reciprocal bistatic sensing channels $\widehat{\mathbf{H}}_{n,m}$ and $\widehat{\mathbf{H}}_{m,n}$ satisfy the offset reciprocity as

Figures (10)

  • Figure 1: Illustration of the D-ISAC system with time and frequency offsets between nodes. Signals from other nodes are omitted for clarity. Each node is equipped with individual LO, resulting in time and frequency offsets between different nodes.
  • Figure 2: (a) Illustration of reciprocal bistatic sensing links utilized for the proposed over-the-air time-frequency synchronization. Relative (b) timing and (c) frequency diagrams between nodes are described with three scatters used for the synchronization.
  • Figure 3: Time-frequency offset estimation performance between two ISAC nodes using reciprocal bistatic sensing channels: (a) TO estimation, and (b) CFO estimation.
  • Figure 4: Impact of the number of OFDM symbols on the proposed SOE approaches: (a) TO estimation, and (b) CFO estimation.
  • Figure 5: Impact of OFDM signal bandwidth on the proposed SOE approaches: (a) TO estimation, and (b) CFO estimation.
  • ...and 5 more figures

Theorems & Definitions (10)

  • Proposition 1
  • proof
  • Proposition 2
  • proof
  • Theorem 1
  • proof
  • Lemma 1
  • proof
  • Theorem 2
  • proof