ISART™ 2024: In Pursuit of Consensus on Clutter

Tutorial

Afternoon of June 10, 2024

ISART Tutorial – Building and Evaluating a Statistical Propagation Model

The purpose of this tutorial is to guide attendees to build their own statistical propagation model based on clutter metrics derived from lidar data. Attendees will build, experiment, and evaluate their propagation models against real clutter measurement datasets. Each attendee will be provided with a cloud-based Python programming environment accessible through the web browsers on their laptops (no software installation necessary). Example code and step-by-step instructions will guide attendees to (a) measure clutter statistics from LiDAR data, (b) formulate a clutter loss model, and (c) assess their model’s generalizability against measurements in new clutter environments. No programming experience necessary, but tutorial attendees must have a computer with a browser and a GitHub account.

For early access to course materials, visit https://github.com/ISART-2024-tutorial/course-materials.

Panel Topics

June 11, 2024

Panel 1 – Model Use Cases and Requirements

Propagation models are used for a wide variety of purposes, from predicting communications link performance, to broadcast coverage, to interference analyses, and more. Each use case drives different propagation model requirements, and the use cases span diverse geometries, including airborne-terrestrial, terrestrial-terrestrial, satellite-terrestrial, etc. Given this broad scope, this panel will describe selected use cases of current and future interest, and discuss propagation model requirements (specifically including clutter) that are derived from their applications. Panelists/speakers will address: What are the different use cases for propagation/clutter models? What is known about the problem? What is needed from the output of the model? What propagation metrics need to be predicted? What role does clutter play in the analysis?

Panel 2 – Modeling

The final output of the development of a clutter model is shaped by the multiple decisions that go into it. Some decisions are straightforward, some are more nuanced, and all are influenced by the scenario the modeling is targeted to support. In all cases, documentation about these decisions is often absent from the final publication of the clutter model itself. The result is models with implicit assumptions that can affect everything from their range of applicability to the underlying uncertainty within the model. This panel discusses how to formalize the model development process for greater transparency and usability, from decision-making steps through model evaluation. All models require some simplifying assumptions: how do we identify and capture the impact of those simplifications? What constitutes sufficient documentation of a model? How can we reduce barriers to improving and expanding existing models? How can we objectively evaluate model performance?

June 12, 2024

Panel 3 – Measurements and Data

The development of clutter models requires high quality, diverse, and verifiable measurement data. This panel will look at how we can establish a scalable measurement framework within which organizations can capture and exchange clutter measurement data across an assorted range of systems. What commonly accepted standards are needed? How can trust be established between diverse datasets? What are the implications of aggregating disparate datasets?

Panel 4 – Using Clutter Models in Interference Analyses/Lessons Learned

Incorporation of clutter models in system-level scenario analysis requires careful attention to both the details of the model itself and the system- and scenario-specific considerations that may require additional tuning or corrections. For example, it may be necessary to incorporate corrections for antenna patterns, clutter depth or angle, or extrapolation of other model parameters to apply the most appropriate clutter model to a scenario that has differences in geometry or other factors from those driving the measurements and assumptions used to develop the original model. This panel discusses how systems, sharing architectures, and studies use clutter model outputs, and how models can be adapted to scenarios that may not share the exact assumptions that were used in their development.

June 13, 2024

Panel 5 – Openness, Collaboration, and Growing the Community

The U.S. Government has ambitious plans to utilize mid-band spectrum in support of both governmental and commercial interests. In order to accomplish such goals, the roster of spectrum engineers must grow. Part of growing the workforce is educational, but part is also lowering the bar to entry to attract attention in an increasingly crowded list of technical fields. This panel will look at how embracing open source and open data can grow the number of people wanting to pursue a career in RF measurement or propagation modeling. How can increased transparency of data and tooling open the doors to more productive collaboration among government, industry, and academia?

Wrap-up Discussion and Next Steps

What are next steps? The panel moderators summarize the most important take-aways from ISART 2024, consider whether community consensus is possible on any well-developed idea or solution, and discuss which areas or ideas warrant further research or stakeholder group involvement.