# Synesthesia of Machines (SoM)-Enhanced ISAC Precoding for Vehicular Networks with Double Dynamics

arXiv: https://arxiv.org/abs/2408.13546 · HTML: https://mlsanity.com/paper/2408.13546/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in RL for energy, networks and operations, one of 15 directions in Deep reinforcement learning methods.
- Builds on 17 of its 48 references that are in the map.
- First posted on arXiv 24 Aug 2024.

## Placed in

- Deep reinforcement learning methods / [RL for energy, networks and operations](https://mlsanity.com/deeprl/rl-for-energy-networks-and-operations/)

## Builds on (most cited first, 17 of 17)

- [Continuous control with deep reinforcement learning](https://mlsanity.com/paper/1509.02971/) (2015)
- [Intelligent Multi-Modal Sensing-Communication Integration: Synesthesia of Machines](https://mlsanity.com/paper/2306.14143/) (2023)
- [Cram'er-Rao Bound Optimization for Joint Radar-Communication Design](https://mlsanity.com/paper/2101.12530/) (2021)
- [Radar-assisted Predictive Beamforming for Vehicular Links: Communication Served by Sensing](https://mlsanity.com/paper/2001.09306/) (2020)
- [Bayesian Predictive Beamforming for Vehicular Networks: A Low-overhead Joint Radar-Communication Approach](https://mlsanity.com/paper/2005.07698/) (2020)
- [Integrated Sensing and Communications Towards Proactive Beamforming in mmWave V2I via Multi-Modal Feature Fusion (MMFF)](https://mlsanity.com/paper/2310.02561/) (2023)
- [Integrated Sensing and Communication with mmWave Massive MIMO: A Compressed Sampling Perspective](https://mlsanity.com/paper/2201.05766/) (2022)
- [Estimating Doubly-Selective Channels for Hybrid mmWave Massive MIMO Systems: A Doubly-Sparse Approach](https://mlsanity.com/paper/1912.08380/) (2019)
- [FRaC: FMCW-Based Joint Radar-Communications System via Index Modulation](https://mlsanity.com/paper/2106.14671/) (2021)
- [High-Efficiency Device Positioning and Location-Aware Communications in Dense 5G Networks](https://mlsanity.com/paper/1608.03775/) (2016)
- [MAJoRCom: A Dual-Function Radar Communication System Using Index Modulation](https://mlsanity.com/paper/1909.04223/) (2019)
- [Location-Aided Beamforming in Mobile Millimeter-Wave Networks](https://mlsanity.com/paper/2205.09887/) (2022)
- [On the Performance Gain of Integrated Sensing and Communications: A Subspace Correlation Perspective](https://mlsanity.com/paper/2211.00434/) (2022)
- [Deep Reinforcement Learning Based on Location-Aware Imitation Environment for RIS-Aided mmWave MIMO Systems](https://mlsanity.com/paper/2205.08788/) (2022)
- [Kalman Filter-based Sensing in Communication Systems with Clock Asynchronism](https://mlsanity.com/paper/2211.04644/) (2022)
- [Integrated Sensing and Communications with MIMO-OTFS](https://mlsanity.com/paper/2306.06361/) (2023)
- [Doubly-Dynamic ISAC Precoding for Vehicular Networks: A Constrained Deep Reinforcement Learning (CDRL) Approach](https://mlsanity.com/paper/2405.14347/) (2024)

## Built on by (most cited first, 2 of 2)

- [Synesthesia of Machines (SoM)-Enhanced Sub-THz ISAC Transmission for Air-Ground Network](https://mlsanity.com/paper/2506.12831/) (2025)
- [Synesthesia of Vehicles: Tactile Data Synthesis from Visual Inputs](https://mlsanity.com/paper/2602.01832/) (2026)

## Abstract (arXiv)

Integrated sensing and communication (ISAC) technology is vital for vehicular networks, yet the time-varying communication channels and rapid movement of targets present significant challenges for real-time precoding design. Traditional optimization-based methods are computationally complex and depend on perfect prior information, which is often unavailable in double-dynamic scenarios. In this paper, we propose a synesthesia of machine (SoM)-enhanced precoding paradigm that leverages modalities such as positioning and channel information to adapt to these dynamics. Utilizing a deep reinforcement learning (DRL) framework, our approach pushes ISAC performance boundaries. We also introduce a parameter-shared actor-critic architecture to accelerate training in complex state and action spaces. Extensive experiments validate the superiority of our method over existing approaches.