[CST]Lower edge frequency of a UWB antenna

i tried to recover the result of one paper using CST in order to do some further work and my focus is the impedance bandwidth of the UWB monopole antenna.
Generally the simulation results is acceptable, while the only inconsistancy with the measured result provided in the paper is the lower edge frequency. In my simulation, it is about 1.98GHz, a little bit higher than the 1.713GHz, which is the measured result shown in the original paper. It is a rather critical value because the UMTS band(1.92-2.17GHz) is supposed to be covered.
Well, i'm fairly confident about my simulation since the adaptive mesh refinement has already been adopted and only 2 passes were executed and stopped normally before the deltaS was satisfied. But i'm wondering whether i could also get the measured result as the paper if the prototype is fabricated.
BTW, now i even suppose the mesh was over-refined for the "lines per wavelength" was chosen to be 30 and finally it took me ~30 hours to complete the simulation.
Could you plz check the results(attached) and help me make the judgement, offer some comment or advise on how to get a better lower edge frequency?
Thank you very much!

there is no any result!
Only structure .mod file!
Can u check again ?

Thank you for your attention!
I reduced the linesperwavelength from 30 to 10, so that the total number of meshcells is dramatically minimized to about only 89000 and the simulation took only a few minutes--2 passes of adaptive mesh refinement were also adopted, rather than the 30+hours before. The S11 results was almost the same but the lower edge frequency dropped to 1.83GHz this time.
I really don't know which results i should believe now.
Could someone help me make a judgement? Thank you!

dear,
please check the feeder of the antenna in the paper that you are mentioned, this may cause the problem. if possible upload the antenna paper that you are referring to.

the match between simulations and measurements really depends on how well you model the structure, i.e. the question is:
Do you really model what you build?
If you have a balun, how well that balun was modeled?

The parameters given in the paper are all crystal clear and i'm sure about the modeling of the structure. There is no need to use a balun since the monopole is fed by a coaxial line.
What i'm interested in now is the influence of the number of the meshcells to the results.

Hi DrunkBear
I have resimulate it and with chage in the mesh and resonance in my case occurs at 1.79 GHZ. Can u please upload the paper.

hi snkhan,
thanks for your reply! The article is given by helio1972 and could you pls tell me the version of your CST and your mesh properties?

any one have any paper about uwb balun design?