Abstract: A method and data structure for preventing the misconcatenation of partial packets when an even number of packets are lost over the network. A channel identifier code is added to the head of a remainder partial packet to allow the remainder partial packet to reassembled with a prior partial packet having the same channel identifier code. The method includes the steps of determining whether each packet will be divided into a first partial packet and a remainder partial packet at an ATM cell boundary, adding a CID at the beginning of each remainder partial packet, assembling the first and remainder partial packets into ATM cells, transmitting the ATM cells, receiving the ATM cells, verifying whether a first partial packet and a remainder partial packet in the received ATM cells have the same CID and assembling the first partial packet and the remainder partial packet into a single packet when the CIDs match.