SymNet: Joint Radio Map Reconstruction and Transmitter Localization
A novel framework named SymNet has been introduced to simultaneously estimate transmitter locations and directional radio maps from limited signal data. This method fills a void left by earlier techniques that primarily concentrate on omnidirectional signals, treating the tasks of transmitter localization and signal map reconstruction as distinct. In omnidirectional scenarios, the highest signal point usually aligns with the transmitter, minimizing the necessity for integrated modeling. However, in directional environments, factors like angular variations, reflections, and building obstructions are crucial, rendering that assumption ineffective. SymNet features a prediction head for locating transmitters while reconstructing radio maps, thus facilitating the concurrent learning of both tasks and utilizing their interrelated information. The findings are published on arXiv (arXiv:2608.00087v1) and are significant for wireless applications needing precise directional radio map predictions.
Key facts
- SymNet is a unified framework for joint directional radio map reconstruction and transmitter localization.
- Prior approaches focus on omnidirectional signals and treat the two tasks separately.
- In omnidirectional settings, the maximum signal location often coincides with the transmitter position.
- Directional propagation involves angular effects, reflections, and building occlusions.
- SymNet incorporates a prediction head for transmitter localization alongside radio map reconstruction.
- The joint formulation leverages complementary information between the tasks.
- The paper is available on arXiv with identifier arXiv:2608.00087.
- The research is relevant to wireless applications requiring accurate directional radio map prediction.
Entities
Institutions
- arXiv