main causes are related to interconnexion of transcoder to MSC (A interface), or transcoder to BSC (Ater interface), as far as I know. But I’m not an expert in the field though.
From BSS QoS point of view, you can’t see any problem. You can just detect that TCH duration is very short, while there is no call drop. It means the subscriber hangs up by himself after just few seconds.
But it’s just a side effect, and if this problem is marginal then the TCH duration in your network is probably not impacted.
To add with the discussion, u hav not mentioned about the exact call case. If your call scenario passes thro several exchanges then u have to verify the whole connections. The main cause would be the physical E1 mismatch or E1 swap. Just make a test call in each E1 and in a particular call you would get the swapped E1s. If this is in a particular cell alone, then the cell parameters have to be checked.