wht the gain of helical antenna simulated by hfss is wrong?

i simulate a gsm dualband helical antenna in hsff. but the peak gain and directivity computed by hfss is always below -10dbi, this is obviously wrong, and the radiated power always large than the accept power, how it can be? But the radition pattern and return loss seem to be correct.

pls help me.

i have the same problem