DATEDIFF

In that case i would advice you to start with creating a categorizing column, converting each datetime to whole and half hour. 

Here I'm rounding up to the nearest whole or half hour. 

2019-01-01 09:49   ->   2019-01-01 10:00
2019-01-01 10:19   ->   2019-01-01 10:30
2019-01-01 10:31   ->   2019-01-01 11:00
2019-01-01 10:54   ->   2019-01-01 11:00

This column you will then be able to use as a grouping column. In other words when you calculate the number of calls you will group the phone number together with this column to create exactly how many intervals are present for the specific caller. 

The measure will look something along the lines of:

Measure = COUNTROWS(SUMMARIZE(FILTER(Table , Table[Status] = "IN"), Table[Mobile] , Table[Calculated Column]))

But it will be easier to work out once the column is created. 

/ J