I was able to migrate the tables and their relationship (one to many) following these steps:

1. Manually created the tables in the Dataverse with the same columns as the tables I wanted to import. I also added the relationship between them (a look up column).
2. I created an empty Access database. I opened it and used the Import from SQL Server option, to take the two tables from SQL Server.
3. In Power Apps, for each of the tables I generated a view and exported to Excel.
4. I exported the data from Access to Excel for each of the tables.
5. For each table, I opened the file with the exported data and copied and pasted it into the Excel template file created in the step 3.
6. For each table, from Power Apps, I open the target table and chose the Import from Excel option, to import the data.
7. To create the relationship, I created a new cloud flow in Power Automate. 8. In The flow, I get the list of records from the look up table, then iterate the list and for each record I take the primary key and assign it to a variable. I also take the GUID for the record (assigned by the Dataverse automatically when record was added in the table upload) and keep it in another variable.
8. I use the variable from step 8, and search in the other table (not the look up table) for the foreign key. For that record, I update the look up column (that defines the link or relationship to the look up table) with the second variable obtained in step 8. I save the record.
9. Once all records have been updated, data is ready and available for use. A check can be done comparing data from source database and data in the Dataverse.