Skip to content

README.md

Latest commit

 

History

History
113 lines (77 loc) · 5.33 KB

README.md

File metadata and controls

113 lines (77 loc) · 5.33 KB

Dataset Overview

This directory contains the datasets used for network modeling and experimentation. Two versions are provided: a Full Dataset and a Small Dataset. Both share the same structure and file types but differ in scale and complexity.

📂 dataset/ — Input datasets for experiments

  • 📁 full-dataset/ — Large-scale dataset for full experiments & performance evaluation
  • 📁 small-dataset/ — Lightweight dataset for debugging & quick tests
  • 📄 README.md — Documentation of dataset format & usage

File Overview

Each dataset contains the following files:

  • centralised_units.json – Centralised Unit (CU) information.
  • distributed_units.json, distributed_units_exist.json, distributed_units_new.json – Distributed Unit (DU) details, distinguishing between existing and newly deployed units.
  • du_cu_path.json, du_cu_path_exist.json, du_cu_path_new.json, ru_du_path.json, ru_du_path_exist.json, ru_du_path_new.json – Path data between network components.
  • radio_units.json, radio_units_exist.json, radio_units_new.json – Radio Unit (RU) details, separated into existing and new devices.
  • region.json – Region definition.
  • road_distances.json, road_edges.json, road_nodes.json – Road network data used for connectivity modeling.
  • user_map.json – Mapping of users to available Radio Units (RUs) and coverage areas.
  • demand_points.json - Contains the grid of demand points for the network, which are used to simulate user demand and coverage requirements.
  • ru_du_existing_mappings.json – Mapping of existing RUs to existing DUs, used for scenarios with pre-deployed infrastructure.

Existing vs New Devices

For both Distributed Units (DUs) and Radio Units (RUs), the dataset separates existing devices from new devices:

  • Existing devices typically carry different cost parameters.
  • New devices model potential deployments.
  • To simulate a greenfield deployment, move all devices from the "existing" files into the "new" files, treating the scenario as if no infrastructure is already in place.

Modifying Dataset Parameters

The JSON files can be edited to explore different network configurations and scenarios:

  • Radio Units (RUs):

    • Coverage radius

    • Capacity bandwidth

    • Location (move, add, or remove RUs)

      If you change the coverage radius or RU locations, you must regenerate the user map to ensure user-to-RU assignments are accurate. Use the helper script make_user_mapping.py for this purpose.

      Note: Adjusting coverage radius or RU placement may result in some coverage targets becoming unattainable. You can modify the required coverage percentage in the config.py file, where all dataset parameters are defined.

  • Distributed Units (DUs):

    • Capacity bandwidth
    • Capacity ports
    • Location (move, add, or remove DUs)

When you add, remove, or relocate RUs or DUs, you also need to rebuild shortest paths for network connectivity. Use the script build_all_paths.py to update all path files accordingly.

  • RU-DU Existing Mapping:
    • The file ru_du_existing_mapping.json defines the mapping between existing RUs and existing DUs for scenarios with pre-deployed infrastructure.
    • If you have multiple existing DUs, each existing RU should be mapped to only one DU in this file. This ensures clear assignment and avoids ambiguity in network topology.
    • Edit ru_du_existing_mapping.json to reflect your desired existing RU-to-DU relationships for different scenarios.

These modifications enable experiments with varying network densities, user demands, and deployment strategies.

Full Dataset

Path: full-dataset/

The full dataset provides comprehensive data for large-scale network modeling. It contains detailed information on all network components, supporting both mixed (existing + new) and greenfield scenarios.

Full Dataset with user grid
Full Dataset with user grid 		Full Dataset
Full Dataset

Small Dataset

Path: small-dataset/

The small dataset is a scaled-down version of the full dataset, designed for:

  • Quick testing
  • Debugging
  • Development and validation runs

It shares the same file structure, including separate files for existing and new devices, but contains fewer entries for faster processing and easier debugging.

Small Dataset
Small Dataset

Usage

  1. Choose the dataset appropriate for your needs:

    • Full Dataset for large-scale, detailed experiments.
    • Small Dataset for rapid iteration or debugging.

  2. Modify parameters within the dataset files to explore different deployment and capacity scenarios.

  3. If changes are made to RU coverage radii, regenerate the user_map.json using the provided script (see above).