Acknowledgment is the procedure which confirms the correct delivery of a piece of information (PDU) to the intended recipient. Acknowledgment procedures may be implemented at any protocol layer, but are particularly important in a protocol layer which provides reliability, such as hop-by-hop acknowledgment in a reliable link protocol or end-to-end acknowledgment in a transport protocol (e.g. TCP) or by an application (e.g. the trvial file transfer protocol, TFTP).
Further detailed information (not included in the EG3567 course)
Acknowledgment may be either:
In a discrete system, the sender asks the receiver which PDUs have been received. This happens only at discrete moments in time (usually on expiry of a timer), and is also known as Check-Pointing (or polling). The receiver responds to each inquiry by returning information from which the sender can then use to determine if any PDUs are required to be retransmitted.
A continuous system allows the receiver to transmit acknowledgement information whenever the protocol determines that this is required. This may be after receipt of each PDU, or may be after a (short) period of time, allowing the acknowledgment for one or more PDUs to be sent in a single reply. In continuous mode, the transmitter does not normally need to send an enquiry, since acknowledgements will normally be received anyway.
Both systems are defined in the High Level Link Control (HDLC) protocol standard, and either or both may be used by an HDLC link protocol. TCP uses a continuous system.