Network Working Group C. Bormann
Request for Comments: 3241 TZI/Uni Bremen
Updates: 1332 April 2002
Category: Standards Track
Robust Header Compression (ROHC) over PPP
Status of this Memo
This document specifies an Internet standards track protocol for the
Internet community, and requests discussion and suggestions for
improvements. Please refer to the current edition of the "Internet
Official Protocol Standards" (STD 1) for the standardization state
and status of this protocol. Distribution of this memo is unlimited.
Copyright (C) The Internet Society (2002). All Rights Reserved.
This document describes an option for negotiating the use of robust
header compression (ROHC) on IP datagrams transmitted over the
Point-to-Point Protocol (PPP). It defines extensions to the PPP
Control Protocols for IPv4 and IPv6.
Robust Header Compression (ROHC) as defined in [RFC3095] may be used
for compression of both IPv4 and IPv6 datagrams or packets
encapsulated with multiple IP headers. The initial version of ROHC
focuses on compression of the packet headers in RTP streams, while
supporting compression of other UDP flows; however, it also defines a
framework into which further header compression mechanisms can be
plugged as new profiles. Planned additions to the set of profiles
supported by ROHC will be capable of compressing TCP transport
protocol headers as well.
In order to establish compression of IP datagrams sent over a PPP
link each end of the link must agree on a set of configuration
parameters for the compression. The process of negotiating link
parameters for network layer protocols is handled in PPP by a family
of network control protocols (NCPs). Since there are separate NCPs
for IPv4 and IPv6, this document defines configuration options to be
used in both NCPs to negotiate parameters for the compression scheme.
ROHC does not require that the link layer be able to indicate the
types of datagrams carried in the link layer frames. However, there
are two basic types of ROHC headers defined in the ROHC framework:
small-CID headers (zero or one bytes are used to identify the
compression context) and large-CID headers (one or two bytes are used
for this purpose). To keep the PPP packets self-describing, in this
document two new types for the PPP Data Link Layer Protocol Field are
defined, one for small-CID ROHC packets and one for large-CID ROHC
packets. (This also avoids a problem that would occur if PPP were to
negotiate which of the formats to use in each of IPCP and IPV6CP and
the two negotiation processes were to arrive at different results.)
A PPP ROHC sender may send packets in either small-CID or large-CID
format at any time, i.e., the LARGE_CIDS parameter from [RFC3095] is
not used. Any PPP ROHC receiver MUST be able to process both small-
CID and large-CID ROHC packets, therefore no negotiation of this
function is required.
ROHC assumes that the link layer delivers packets in sequence. PPP
normally does not reorder packets. When using reordering mechanisms
such as multiclass multilink PPP [RFC2686], care must be taken so
that packets that share the same compression context are not
reordered. (Note that in certain cases, reordering may be acceptable
to ROHC, such as within a sequence of packets that all do not change
the decompression context.)
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in RFC 2119.
2. Configuration Option
This document specifies a new compression protocol value for the IPCP
IP-Compression-Protocol option as specified in [RFC1332]. The new
value and the associated option format are described in section 2.1.
The option format is structured to allow future extensions to the
It may be worth repeating [RFC1332], section 4: "The IP-Compression-
Protocol Configuration Option is used to indicate the ability to
receive compressed packets. Each end of the link must separately
request this option if bi-directional compression is desired." I.e.,
the option describes the capabilities of the decompressor (receiving
side) of the peer that sends the Configure-Request.
NOTE: The specification of link and network layer parameter
negotiation for PPP [RFC1661], [RFC1331], [RFC1332] does not
prohibit multiple instances of one configuration option but states
that the specification of a configuration option must explicitly
allow multiple instances. From the current specification of the
IPCP IP-Compression-Protocol configuration option [RFC1332] one
can infer that it can only be used to select a single compression
protocol at any time.
This was appropriate at a time when only one header compression
scheme existed. With the advent of IP header compression
[RFC2507, RFC2509], this did not really change, as RFC 2507
essentially superseded RFC 1144. However, with ROHC, it may now
very well be desirable to use RFC 2507 TCP compression in
conjunction with RFC 3095 RTP/UDP compression.
The present document now updates RFC 1332 by explicitly allowing the
sending of multiple instances of the IP-Compression-Protocol
configuration option, each with a different value for IP-
Compression-Protocol. Each type of compression protocol may
independently establish its own parameters.
This change is believed to not cause significant harm in existing PPP
implementations, as they would most likely Configure-Nak or
Configure-Reject the duplicate option, or simply happen to accept the
one option they understand. To aid interoperability, the peer
implementing the present specification SHOULD react to a Configure-
Nak or Configure-Reject by reducing the number of options offered to
2.1. Configuration Option Format
Both the network control protocol for IPv4, IPCP [RFC1332] and the
IPv6 NCP, IPV6CP [RFC2472] may be used to negotiate IP Header
Compression parameters for their respective protocols. The format of
the configuration option is the same for both IPCP and IPV6CP.
This NCP configuration option is used to negotiate parameters for
Robust Header Compression. The option format is summarized below.
The fields are transmitted from left to right.
Figure 1: Robust Header Compression (ROHC) Option
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
| Type | Length | IP-Compression-Protocol |
| MAX_CID | MRRU |
| MAX_HEADER | suboptions...
The length may be increased if the presence of additional
parameters is indicated by additional suboptions.
The MAX_CID field is two octets and indicates the maximum value of
a context identifier.
Suggested value: 15
MAX_CID must be at least 0 and at most 16383 (The value 0 implies
having one context).
The MRRU field is two octets and indicates the maximum
reconstructed reception unit (see [RFC3095], section 5.1.1).
Suggested value: 0
The largest header size in octets that may be compressed.
Suggested value: 168 octets
The value of MAX_HEADER should be large enough so that at least
the outer network layer header can be compressed. To increase
compression efficiency MAX_HEADER should be set to a value large
enough to cover common combinations of network and transport layer
NOTE: The four ROHC profiles defined in RFC 3095 do not provide
for a MAX_HEADER parameter. The parameter MAX_HEADER defined by
this document is therefore without consequence in these profiles.
Other profiles (e.g., ones based on RFC 2507) can make use of the
parameter by explicitly referencing it.
The suboptions field consists of zero or more suboptions. Each
suboption consists of a type field, a length field and zero or
more parameter octets, as defined by the suboption type. The
value of the length field indicates the length of the suboption in
its entirety, including the lengths of the type and length fields.
Figure 2: Suboption
0 1 2
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3
| Type | Length | Parameters...
2.2. PROFILES Suboption
The set of profiles to be enabled is subject to negotiation. Most
initial implementations of ROHC implement profiles 0x0000 to 0x0003.
This option MUST be supplied.
Define the set of profiles supported by the decompressor.
Figure 3: PROFILES suboption
0 1 2
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3
| Type | Length | Profiles...
n octet-pairs in ascending order, each octet-pair specifying a
ROHC profile supported.
3. Multiple Network Control Protocols
The ROHC protocol is able to compress both IPv6 and IPv4 datagrams.
Both IPCP and IPV6CP are able to negotiate option parameter values
for ROHC. The ROHC capability negotiated as a whole applies to the
compression of packets where the outer header is an IPv4 header and
an IPv6 header, respectively; e.g., an outer IPv6 header MUST NOT be
sent if the ROHC IP-Compression-Protocol option was not negotiated
Offering a specific ROHC capability in a Configure-Request in either
IPCP or IPV6CP indicates that the capability is provided for the
entire ROHC channel formed by the PPP link. When the option has been
negotiated with different values in IPCP and IPV6CP, the result is
that the set of parameter values for the entire ROHC channel is the
logical union of the two values, i.e., the maximum for MAX_CID, MRRU
or MAX_HEADER, and the logical union of the suboptions. For the
PROFILES suboption, the logical union is the union of the two sets of
profiles. The unified values are kept as valid parameter values for
the ROHC channel even when either of the NCPs is taken down.
Note that each new suboption for this option must define the meaning
of "logical union", if the concept applies.
3.1. Sharing Context Identifier Space
For the compression and decompression of IPv4 and IPv6 datagram
headers, the context identifier space is shared. While the parameter
values are independently negotiated, sharing the context identifier
spaces becomes more complex when the parameter values differ. Since
the compressed packets share context identifier space, the
compression engine must allocate context identifiers out of a common
pool; for compressed packets, the decompressor has to examine the
context state to determine what parameters to use for decompression.
In particular, the context identifier space is shared between ROHC
small-CID packets and ROHC large-CID packets. From the point of view
of the ROHC framework, the PPP NCP instances for IPCP and IPV6CP
together constitute exactly one ROHC channel; its feedback is
destined for the ROHC channel defined by the NCP instances for IPCP
and IPV6CP in the reverse direction on the same PPP link.
In particular, this means that taking down either of the NCPs while
the other is still open means that the contexts of the channel stay
active. To avoid race conditions, the same is true if both NCPs are
taken down and then one or more is reopened. Taking down LCP
destroys the channel, however; reopening LCP and then one or more of
IPCP and IPV6CP restarts ROHC with all contexts in no-context state.
4. Demultiplexing of Datagrams
The ROHC specification [RFC3095] defines a single header format for
all different types of compressed headers, with a variant for small
CIDs and a variant for large CIDs. Two PPP Data Link Layer Protocol
Field values are specified below.
The frame contains a ROHC packet with small CIDs as defined in
Value: 0003 (hex)
The frame contains a ROHC packet with large CIDs as defined in
Value: 0005 (hex)
Note that this implies that all CIDs within one ROHC packet MUST be
of the same size as indicated by the Data Link Layer Protocol field,
either small or large. In particular, embedded feedback MUST have a
CID of the same size as indicated by the Protocol field value. For
piggybacking feedback, a compressor must be able to control the
feedback CID size used by the associated decompressor, ensure that
all CIDs are of the same size, and indicate this size with the
appropriate Protocol Field value.
To make CID interpretation unambiguous when ROHC segmentation is
used, all packets that contribute to a segment MUST be sent with the
same Data Link Layer Protocol Field value, either 0003 or 0005, which
then also applies to the CID size in the reconstructed unit. A unit
reconstructed out of packets with Protocol field values that differ
MUST be discarded.
5. ROHC Usage Considerations
Certain considerations are required for any ROHC-over-X protocol.
This section describes how some of these are handled for ROHC over
5.1. Uncompressed profile
There is no need for the ROHC uncompressed profile in ROHC over PPP,
as uncompressed packets can always be sent using the PPP protocol
demultiplexing method. Therefore, no consideration was given to
locking down one of the context numbers for the uncompressed profile
(see [RFC3095] section 5.1.2). Note, however, that according to the
ROHC specification, profile 0x0000 must not be rejected [RFC3095], so
it MUST be implemented by all receivers.
5.2. Parameter selection
For each of the ROHC channel parameters MAX_CID and MRRU, the value
is the maximum of the respective values negotiated for the IPCP and
IPv6CP instances, if any. The ROHC channel parameter FEEDBACK_FOR is
set implicitly to the reverse direction on the same PPP link (see
"Sharing Context Identifier Space" above). The ROHC channel
parameter LARGE_CIDS is not used, instead the PPP protocol ID on the
packet is used (see "Demultiplexing of Datagrams" above).
A number of parameters for ROHC must be set correctly for good
compression on a specific link. E.g., the parameters k_1, n_1, k_2,
n_2 in section 184.108.40.206.3 of [RFC3095] need to be set based on the
error characteristics of the underlying links. As PPP links are
usually run with a strong error detection scheme [RFC1662], k_1 = n_1
= k_2 = n_2 = 1 is usually a good set of values. (Note that in any
case k values need to be set low enough relative to n values to allow
for the limited ability of the CRC to detect errors, i.e., the CRC
will succeed for about 1/8 of the packets even in case of context
damage, so k/n should be significantly less than 7/8.)
6. Security Considerations
Negotiation of the option defined here imposes no additional security
considerations beyond those that otherwise apply to PPP [RFC1661].
The security considerations of ROHC [RFC3095] apply.
The use of header compression can, in rare cases, cause the
misdelivery of packets. If necessary, confidentiality of packet
contents should be assured by encryption.
Encryption applied at the IP layer (e.g., using IPSEC mechanisms)
precludes header compression of the encrypted headers, though
compression of the outer IP header and authentication/security
headers is still possible as described in [RFC3095]. For RTP
packets, full header compression is possible if the RTP payload is
encrypted by itself without encrypting the UDP or RTP headers, as
described in [RFC1889]. This method is appropriate when the UDP and
RTP header information need not be kept confidential.
7. IANA considerations
The ROHC suboption identifier is a non-negative integer. Following
the policies outlined in [RFC2434], the IANA policy for assigning new
values for the suboption identifier shall be Specification Required:
values and their meanings must be documented in an RFC or in some
other permanent and readily available reference, in sufficient detail
that interoperability between independent implementations is
possible. The range 0 to 127 is reserved for IETF standard-track
specifications; the range 128 to 254 is available for other
specifications that meet this requirement (such as Informational
RFCs). The value 255 is reserved for future extensibility of the
The following suboption identifiers are already allocated:
Suboption Document Usage
1 RFC3241 Profiles
The RFC 3006 compressibility hint [RFC3006] for ROHC is 0x0003pppp,
where 0xpppp is the profile assumed.
(Note that the PPP protocol identifier values 0003 and 0005 were
taken from a previously reserved space that exhibits inefficient
transparency in the presence of asynchronous control character
escaping, as it is considered rather unlikely that ROHC will be used
over links with highly populated ACCMs.)
The present document borrows heavily from [RFC2509].
The author would like to thank Pete McCann and James Carlson for
clarifying the multiple option instance issue, Craig Fox for helping
with some PPP arcana, and Lars-Erik Jonsson for supplying some final
9.1. Normative References
[RFC1332] McGregor, G., "The PPP Internet Protocol Control Protocol
(IPCP)", RFC 1332, May 1992.
[RFC1661] Simpson, W., Ed., "The Point-To-Point Protocol (PPP)", STD
51, RFC 1661, July 1994.
[RFC2472] Haskin, E. and E. Allan, "IP Version 6 over PPP", RFC 2472,
[RFC3006] Davie, B., Casner, S., Iturralde, C., Oran, D. and J.
Wroclawski, "Integrated Services in the Presence of
Compressible Flows", RFC 3006, November 2000.
[RFC3095] Bormann, C., Burmeister, C., Degermark, M., Fukushima, H.,
Hannu, H., Jonsson, L-E., Hakenberg, R., Koren, T., Le, K.,
Liu, Z., Martensson, A., Miyazaki, A., Svanbro, K., Wiebke,
T., Yoshimura, T. and H. Zheng, "RObust Header Compression
(ROHC): Framework and four profiles: RTP, UDP, ESP, and
uncompressed", RFC 3095, July 2001.
9.2. Informative References
[RFC1144] Jacobson, V., "Compressing TCP/IP Headers for Low-Speed
Serial Links", RFC 1144, February 1990.
[RFC1889] Schulzrinne, H., Casner, S., Frederick, R. and V.
Jacobson, "RTP: A Transport Protocol for real-time
applications", RFC 1889, January 1996.
[RFC2434] Alvestrand, H. and T. Narten, "Guidelines for Writing an
IANA Considerations Section in RFCs", BCP 26, RFC 2434,
[RFC2507] Degermark, M., Nordgren, B. and S. Pink, "IP Header
Compression", RFC 2507, February 1999.
[RFC2509] Engan, M., Casner, S. and C. Bormann, "IP Header
Compression over PPP", RFC 2509, February 1999.
[RFC2686] Bormann, C., "The Multi-Class Extension to Multi-Link PPP",
RFC 2686, September 1999.
11. Full Copyright Statement
Copyright (C) The Internet Society (2002). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assigns.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Funding for the RFC Editor function is currently provided by the