I want to propose some approaches in wireless networks and I want to simulate it to prove that it is better than others; but I don't want use commonly-used simulators and I want to create new simulator according to requirements. Is there any difference in paper acceptance for international conferences to use your own simulator or famous simulators?
The important thing is to use the right tool for the job, and justify it.
The existing simulators have the following advantage: a great deal of time and effort went into their development, and they have been used, debugged, and extended by many people. Therefore, they are likely to be more complete and/or more accurate than anything you could come up with on your own in a reasonable amount of time.
However, if you have a compelling reason why these are not suitable for the particular bit of research you want to do, by all means write your own simulator; but you should be able to justify this in your paper.
For some objective numbers, I refer you to this survey of papers published in MobiHoc from 2000-2005. 80 of the 114 papers that used simulations stated what simulator they used. From those, 22/80 said they used "self-developed or custom simulators" (27.3%).