{"draft":"draft-ietf-avt-hc-mpls-reqs-03","doc_id":"RFC4247","title":"Requirements for Header Compression over MPLS","authors":["J. Ash","B. Goode","J. Hand","R. Zhang"],"format":["ASCII","HTML"],"page_count":"11","pub_status":"INFORMATIONAL","status":"INFORMATIONAL","source":"Audio\/Video Transport","abstract":"Voice over IP (VoIP) typically uses the encapsulation voice\/RTP\/UDP\/IP.  When MPLS\r\nlabels are added, this becomes voice\/RTP\/UDP\/IP\/MPLS-labels.  For an\r\nMPLS VPN, the packet header is typically 48 bytes, while the voice\r\npayload is often no more than 30 bytes, for example.  Header compression can\r\nsignificantly reduce the overhead through various compression\r\nmechanisms, such as enhanced compressed RTP (ECRTP) and robust header\r\ncompression (ROHC).  We consider using MPLS to route compressed\r\npackets over an MPLS Label Switched Path (LSP) without compression\/decompression cycles at\r\neach router.  This approach can increase the bandwidth efficiency as\r\nwell as processing scalability of the maximum number of simultaneous\r\nflows that use header compression at each router.  In this document, we\r\ngive a problem statement, goals and requirements, and an example\r\nscenario.  This memo provides information for the Internet community.","pub_date":"October 2005","keywords":["multiprotocol label switching","voip","voice over ip"],"obsoletes":[],"obsoleted_by":[],"updates":[],"updated_by":[],"see_also":[],"doi":"10.17487\/RFC4247","errata_url":null}