Settings of the IRT database in WallMan for the V2X

Mateusz_22418

Dear Sir/Madam,

I am using your WinProp software, version 2023.1.2, for my research project titled "Analysis of Vehicle-to-Vehicle and Vehicle-to-Infrastructure Communication in an Urban Environment." The area for which I am defining communication is approximately 270 x 270 meters. The main communication occurs within a single intersection, with street widths ranging from about 10 to 15 meters. The vehicle-simulating objects used in the project vary in size, with dimensions ranging from 1.8/4.5/1.8 meters (height/length/width) to 4.1/15/2.55 meters. I am specifically analyzing a blockage scenario.

I have a few questions regarding the pre-processing settings for the IRT model.

For such a large database—where time-varying simulation requires the use of an internal database—the default parameter settings, such as:

- Minimum wedge length: 0.01 m
- Panel width for standard walls: 2 m;
- Panel width on predic: 2 m
- Wedge segment length: 2 m
- Width of tiles at predicion planes: 2 m
- Spherical zone (disabled);
- Exclusion of triangular objects (disabled); 
generates a database of approximately 14 GB for just one second of simulation, and I would like to consider at least 15 seconds for the scenario.

Therefore, I would like to reduce the database size significantly. What settings should I use to ensure that the approximation generated during the simulation remains as close as possible to the real-world scenario, while keeping the database size much smaller?
Best regards,
Mateusz

Zeina

Hi Mateusz,

I assume that you are preprocessing an indoor database that actually models an urban environnement.

There are multiple factors that would affect the size of your preprocessed-database in this case:

• The original size of your .idb: Since you are modeling an urban environnement as an indoor database, the number of objects is much higher than an outdoor database. This might already be a prohibitive size.
• The number of time-steps you are considering in your preprocessing: The more steps you include, the bigger the preprocessed database.
• The parameters you have mentioned above for your preprocessing: I don't think this will have a big impact on the size of the database. You can try to increase the minimum wedge length for instance, this could include less wedges. But for the other four settings, since you have objects <2m, I don't think it would be wise to increase the rest.

I would suggest that you reduce your indoor database to the intersection of interest and remove the other parts that seem to have not much impact on the communication based on your description.

I hope this helps!

Zeina

Mateusz_22418

Hello Zeina,

First of all, I would like to sincerely thank you for your quick response.

You are right; I am processing an indoor database that indeed models an urban environment. The source database I am processing is about 2.6 MB in size. The number of walls for the processed area is 11,770, with the majority being walls representing vehicle edges. Regarding the number of steps, I decided to divide the scenario into smaller fragments, as the amount of RAM I have on my workstation is insufficient (I am encountering error 901 related to memory allocation).

As for the parameters, I wanted to highlight that changing:

the minimum length of Wedges to 1m and all subdivisions to 5m reduced the database size by 50% (I am modeling an urban canyon environment).

However, I would like to ask how much this would affect the quality of the results. I found information in the documentation stating that only walls larger than the specified tile subdivision value will be divided, while smaller ones will remain unchanged. However, I want to confirm that the smaller walls will indeed remain in the database. This is important to me because, in my case, the smaller walls represent moving vehicles, and the phenomena observed on them are crucial in my scenario involving a blockage.

Additionally, I wanted to ask about the maximum inner angle of wedges. Is setting it to the default 180 degrees too large a value? It seems that 90 degrees might be more appropriate, or am I mistaken?

In my project, besides propagation analysis, I also intend to study throughput depending on the communication protocol. Therefore, I need to cover an area that allows me to investigate not only path loss but also throughput.

A final question is how to configure the base station to act as a repeater.

I apologize for any mistakes in my English.

Best regards,
Mateusz

Zeina

Hi Mateusz,

You're welcome!

I see that your initial database size is relatively reasonable so I wonder why not solve it directly using the Standard Ray Tracing (SRT) as a Time-Varying scenario in ProMan, without performing a preprocessing? I would expect a good runtime.

Best,

Zeina