To provide a possibility to have a near-real-time replay of a pre-empted communication while ensuring that the pre-empting communication is not interrupted, when it is detected that a pre-empting communication is pre-empting a pre-empted communication, the pre-empted communication is buffered to a buffer for a replay, and a safety interval for the pre-empting communication is determined. The pre-empting communication is determined to be ended in response to an expiry of the safety interval after end of the pre-empting communication.