RoLocMe: A Robust Multi-agent Source Localization System with Learning-based Map Estimation

Authors: Thanh Dat Le, Lyuzhou Ye, Yan Huang

IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental This section presents experiments conducted to evaluate the Ro Loc Me system in comparison with the state-of-arts methods and provides an in-depth analysis of the significance of each component in its design. Section 4: Experiments and Results. Figure 6: Performance benchmark of 4 agents of Ro Loc Me with other RL models in the testing set. Section 4.5: Drop-One Component Ablation Study.
Researcher Affiliation Academia Thanh Dat Le , Lyuzhou Ye , Yan Huang University of North Texas EMAIL, EMAIL, EMAIL
Pseudocode Yes Algorithm 1 details the entire procedure. Algorithm 1 Ro Loc Me Training
Open Source Code No The paper does not contain any explicit statement about releasing source code for the methodology described, nor does it provide a link to a code repository.
Open Datasets Yes We use the Radio Map Seer dataset [Yapar et al., 2022]
Dataset Splits Yes We divide the dataset by environment, allocating 501 maps for training, 100 for validation, and 100 for testing.
Hardware Specification No The paper mentions "end-to-end GPU training" but does not specify any particular GPU model, CPU model, or other hardware specifications. It only uses a vague term without specific details.
Software Dependencies No The paper mentions optimizers like RAdam and Adam, citing their original papers, but it does not specify any software libraries (e.g., PyTorch, TensorFlow) with version numbers, nor the programming language version used for implementation.
Experiment Setup Yes Table 1: Hyperparameters used for training Ro Loc Me and Skip Net. This table includes specific values for Learning rate (5e-4), Discount factor (γ) (0.99), Decay rate (λ) (0.65), Number of parallel environments (Nenv) (128), Mini batch size (16), and other parameters.