DInSAR analysis

I use three SLC data as below for DinSAR analysis.

A:S1A_IW_SLC__1SSV_20150418T204258_20150418T204325_005543_007169_C251
B:S1A_IW_SLC__1SSV_20150512T204259_20150512T204326_005893_00796F_7DE0
C:S1A_IW_SLC__1SSV_20150605T204301_20150605T204328_006243_0082A3_A8B3

The result of A&B and B&C shows reverse fringe pattern.

According to GNSS data, this is impossibe and the result shows same fringe pattern.

Is it bag ?

Did you check the all SLC data before processing. Why did you chose B SLC as a master? You should use stack overview, it may help to you.