meshes in the antenna

Hello, are you pointing to actual (physical) mesh structure (for example a reflecting screen) or meshing as used in EM simulation software?

dear sir/madam,
i ask meshing as used in EM simulation software

Like in digital signal processing, the EM problem has to be converted to a discrete problem before it can be solved/processed.

In the time domain, the discretization (sampling rate) determines how fast a time varying signal can be processed. When it changes faster than the sample rate, the processing will not take it into account.

In the space domain, the discretization (meshing) determines how fast a current or voltage amplitude can change over distance.

If you expect a significant current amplitude ir phase change in an antenna structure over a distance of 10 cm, the meshing size should be less then 10 cm, otherwise the solver will not show you the correct current distribution.

Most MoM solvers use default settings for meshing based on the highest frequency and dielectric or magnetic environment. They assume that current and voltage can change significantly over a distance of 0.25 lambda (think of a standing wave transmission line problem). Values of 15..40 segments/lambda are used.

You always have to know your structure, as (for example in structures where conductors are closely spaced) current density may change with distance very fast. Think of a coplanar wave-guide. At the edges facing each other the current density is significantly higher then at some distance from it.

For large problems you have to choose between accuracy and speed. If you have an antenna over a limited ground plane, meshing the ground with same density as the antenna itself may result in very looooong runtime. Using less dense meshing in areas where current changes slowly with distance, virtually doesn't affect accuracy but does reduces runtime significantly.