Internet DRAFT - draft-jiang-semantic-prefix
draft-jiang-semantic-prefix
Network Working Group S. Jiang, Ed.
Internet-Draft Huawei Technologies Co., Ltd
Intended status: Informational Q. Sun
Expires: January 16, 2014 China Telecom
I. Farrer
Deutsche Telekom AG
Y. Bo
Huawei Technologies Co., Ltd
T. Yang
China Mobile
July 15, 2013
Analysis of Semantic Embedded IPv6 Address Schemas
draft-jiang-semantic-prefix-06
Abstract
This informational document discusses the use of embedded semantics
within IPv6 address schemas. Network operators who have large IPv6
address space may choose to embed some semantics into their IPv6
addressing by assigning additional significance to specific bits
within the prefix. By embedding semantics into IPv6 prefixes, the
semantics of packets can be easily inspected. This can simplify the
packet differentiation process. However, semantic embedded IPv6
address schemas have their own operational cost and even potential
pitfalls. Some complex semantic embedded IPv6 address schemas may
also require new technologies in addition to existing Internet
protocols.
The document aims to understand the usage of semantic embedded IPv6
address schemas, and neutrally analyze on the associated advantages,
drawbacks and technical gaps for more complex address schemas.
Status of This Memo
This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet-
Drafts is at http://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
Jiang, et al. Expires January 16, 2014 [Page 1]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
This Internet-Draft will expire on January 16, 2014.
Copyright Notice
Copyright (c) 2013 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document. Code Components extracted from this document must
include Simplified BSD License text as described in Section 4.e of
the Trust Legal Provisions and are provided without warranty as
described in the Simplified BSD License.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 4
3. Understanding of Semantic IPv6 Prefix Address Schema . . . . 4
3.1. Overview of Semantic IPv6 Prefix Address Schema . . . . . 4
3.2. Existing Approaches to Traffic Differentiation . . . . . 5
3.3. Justification for Semantics with the IPv6 Prefix . . . . 6
3.4. The Semantic Prefix Domain . . . . . . . . . . . . . . . 7
3.5. The Embedded Semantics . . . . . . . . . . . . . . . . . 8
3.6. Network Operations Based on Semantic Prefixes . . . . . . 8
4. Potential Benefits . . . . . . . . . . . . . . . . . . . . . 9
5. Potential Drawbacks . . . . . . . . . . . . . . . . . . . . . 10
6. Gaps for complex semantic prefix scenarios . . . . . . . . . 11
6.1. Semantic Notification in the Network . . . . . . . . . . 11
6.2. Semantic Relevant Interactions between Hosts and the
Network . . . . . . . . . . . . . . . . . . . . . . . . . 12
6.3. Additional Technical Extensions . . . . . . . . . . . . . 12
7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 13
8. Change Log (removed by RFC editor) . . . . . . . . . . . . . 13
9. Security Considerations . . . . . . . . . . . . . . . . . . . 14
10. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 14
11. References . . . . . . . . . . . . . . . . . . . . . . . . . 14
11.1. Normative References . . . . . . . . . . . . . . . . . . 15
11.2. Informative References . . . . . . . . . . . . . . . . . 15
Appendix A. An ISP Semantic Prefix Example . . . . . . . . . . . 15
A.1. Function Type Semantic Bits . . . . . . . . . . . . . . . 16
A.2. Network Device Type Bits within Network Device Address
Space . . . . . . . . . . . . . . . . . . . . . . . . . . 17
A.3. Subscriber Type Bits within Subscriber Address Space . . 17
A.4. Service Platform Type Bits within Service Platform
Jiang, et al. Expires January 16, 2014 [Page 2]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
Address Space . . . . . . . . . . . . . . . . . . . . . . 18
Appendix B. An Enterprise Semantic Prefix example . . . . . . . 19
Appendix C. A Multi-Prefix Semantic example . . . . . . . . . . 20
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 21
1. Introduction
As the global Internet expands, it is being used for an increasingly
diverse range of services. These services place differentiated
requirements upon packet delivery networks meaning that Internet
Service Providers and enterprises need to be aware of more
information about each packet in order to best meet a specific
service's needs. Dividing a network into different subnets according
to different semantics is already widely existing today, mostly
motivated by either topological aspects, logical user/device groups,
and/or trust/security domains.
In order to inspect the semantics of packets so that they can be
treated differently, some network operators have chosen to embed
semantics into IPv6 prefixes. Routers and other intermediary devices
can easily apply relevant policies as required. User types, service
types, applications, security requirements, traffic identity types,
quality requirements and other criteria may be used according to how
a network operator may want to differentiate its services. Packet-
level differentiation can also enable flow-level and user-level
differentiation. Consequently, the network operators can treat
network packets differently and efficiently. It is believed this
mechanism can simplify the management and maintenance of networks.
However, semantic embedded IPv6 address schemas come with their own
operational cost and even pitfalls. Some complex semantic embedded
IPv6 address schemas may also require technologies additional to
existing Internet protocols.
While network operators, who already have large IPv6 address space
allocations, are free to plan and deploy addressing in their
preferred way (including semantic embedded IPv6 address schemas), it
is useful to analyze the benefits and drawbacks of a semantic
approach to addressing.
The document only discusses the usage of semantics within a single
network, or group of interconnected networks which share a common
addressing policy, referred to as a Semantic Prefix Domain.
Jiang, et al. Expires January 16, 2014 [Page 3]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
This document does not intend to suggest the standardization of any
common global semantics. It does not intend to draw any conclusions,
either recommending this kind of address schemas or not. It aims to
provide network operators with relevant information to use in the
creation of their own addressing policy.
2. Terminology
The following terms are used throughout this document:
Semantic Prefix: A flexible-length IPv6 prefix which embeds certain
semantics.
Semantic Prefix Domain: A portion of the Internet over which a
consistent semantic-prefix based policy is in operation.
Semantic Prefix Policy: A policy based on the embedded semantics
within IPv6 prefix.
3. Understanding of Semantic IPv6 Prefix Address Schema
Some network operators (either ISPs or enterprise network operators),
who have large IPv6 address space, have chosen to embed certain pre-
defined semantics into their IPv6 address schemas by assigning
additional significance to specific bits within the prefix. The IPv6
addresses of each packet can then explicitly express semantics.
Consequently, intermediate devices can easily apply relevant packet
differentiating operations accordingly. This mechanism may divert
much network complexity to the planning and management of IPv6
addressing and IP address based policies.
For illustrations of how semantic prefixes could be applied in real-
world scenarios, Appendix A describes an ISP example semantic IPv6
prefix address schema; Appendix B introduces an enterprise semantic
IPv6 prefix example; and Appendix C introduces an enterprise example
in which a multiple-site enterprise network with several prefixes of
different lengths is organized as a single, contiguous Semantic
Prefix Domain.
3.1. Overview of Semantic IPv6 Prefix Address Schema
Jiang, et al. Expires January 16, 2014 [Page 4]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
A network operator first plans their IPv6 address schema, in which
useful semantics (see Section 3.5) are embedded into prefix. They
then delegate prefixes with the corresponding semantics to users.
The users generate their IPv6 addresses based on assigned prefixes.
Then, when the IPv6 stack on the user devices forms packets, the
source addresses comprise compliance semantics. For trust reasons,
the filters on the edge router may drop packets which are not
compliant with assigned prefixes.
The embedded semantics are only meaningful within a network domain
which implements a single policy (see Section 3.4). Different
service providers may make very different choices regarding the
specific semantics which are relevant to their networks. Therefore,
it is not possible or even desirable to attempt to standardize a
general semantic prefix policy.
Forwarding policies, access control lists, policy-based routing,
security isolation and other network operations (see Section 3.6) can
be easily applied according to semantics, which are self-expressed by
the source address of every packet. Also, the semantics of the
destination address may be taken in account if the destination is in
the same Semantic Prefix Domain or the peer Semantic Prefix Domain
whose semantics has been notified.
3.2. Existing Approaches to Traffic Differentiation
There are several existing approaches which have been developed that
can assist operators in identifying and marking traffic. These
solutions were mainly developed in the IPv4 era, where the IP address
is used as a host locator and little else. The limited capacity of a
32-bit IPv4 address provides very little room for encoding additional
information. Correspondingly, these approaches are indirect,
inefficient and expensive for operators.
3.2.1. Differentiated Services
Quality of Service (QoS) based on and Differentiated Services
[RFC2474] is a widely deployed framework specifying a simple and
scalable coarse-grained mechanism for classifying and managing
network traffic. But in a service provider's network, DiffServ
codepoint (DSCP) values cannot be trusted when they are set by the
customer as these are arbitrary values.
In real-world scenarios, ISPs deploy "remarking" points at the
customer edge of their network, re-classifying received packets by
rewriting the DSCP field according to local policy using information
such as the source/destination address, IP protocol number and
transport layer source/destination ports.
Jiang, et al. Expires January 16, 2014 [Page 5]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
The traffic classification process leads to increased packet
processing overhead and complexity at the edge of the service
provider's network.
DSCP mechanism abstracts all the semantics into a single-dimension
service classes. This abstract processing has lost a lot of semantic
information, which providers want to inspect for every packet, then
process the packet accordingly.
The DSCP in the IPv6 header traffic class field allows 6-bits for
encoding service provider specific information related to the
contents of the packet. Whilst this is a useful part of an overall
packet differentiation architecture, the relative small number of
available bits (when compared to the available number of bits within
the service providers prefix) means that it cannot be used in
isolation.
3.2.2. Deep Packet Inspection
Deep Packet Inspection (DPI) may also be used by ISPs to learn the
characteristics of users packets. This involves looking into the
packet well beyond the network-layer header to identify the specific
application traffic type. Once identified, the traffic type can be
used as an input for setting the packet's DSCP or other actions.
But DPI is expensive both in processing costs and latency. The
processing costs means that dedicated infrastructure is necessary to
carry out the function. The incurred latency may be too much for use
with any delay/jitter sensitive applications. As a result, DPI is
difficult for large-scale deployment and it's usage is usually
limited to small and specific functions in the network. In short, it
is not scalable, and cannot support realtime network operations.
3.3. Justification for Semantics with the IPv6 Prefix
Although the interface identifier portion of an IPv6 address has
arbitrary bits and extension headers can carry significantly more
information, these fields can not be trusted by network operators.
Users may easily change the setting of interface identifier or
extension headers in order to obtain undeserved priorities/
privileges, while servers or enterprise users may be much more self-
restricted since they are charged accordingly.
With proper access control filters deployed, the prefix can be
trusted by the network operators and is simple to inspect in the IP
header of a packet. The packets with the noncompliance source
addresses should be filtered. The prefix is delegated by the network
and therefore the network is able to detect any undesired
Jiang, et al. Expires January 16, 2014 [Page 6]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
modifications and filter the packet accordingly. This also makes it
possible for the service provider to increase the level of trust in a
customer-generated packet. If the packet has an source or
destination address which is outside of the network operator's policy
then a session will simply fail to establish.
3.4. The Semantic Prefix Domain
A Semantic Prefix Domain is a portion of the Internet over which a
consistent set of semantic-prefix-based policies are administered in
a coordinated fashion. It is analogous to a Differentiated Services
Domain [RFC2474]. Some of the characteristics that a single Semantic
Prefix Domain could represent include:
a. Administrative domains
b. Autonomous systems
c. Trust regions
d. Network technologies
e. Hosts
f. Routers
g. User groups
h. Services
i. Traffic groups
j. Applications
A Semantic Prefix Domain has a set of pre-defined semantic
definitions, which are only meaningful locally. Without an efficient
semantics notification, exchanging mechanism or service agreement,
the definitions of semantics are only meaningful within local
Semantic Prefix Domain. Agreements on definitions between network
operators could be made. However, this may involve trust models
among network operators. Sharing semantic definition among Semantic
Prefix Domains enables more semantic based network operations.
An enterprise Semantic Prefix Domain may span several physical
networks and traverse ISP networks. However, when an interim network
is traversed (such as when an intermediary ISP is used for
interconnectivity), the relevance of the semantics is limited to
network domains that share a common Semantic Prefix Policy.
Jiang, et al. Expires January 16, 2014 [Page 7]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
If an ISP has several non-contiguous address blocks, they may be
organized as a single Semantic Prefix Domain if the same Semantic
Prefix Policy is shared across these non-contiguous address blocks.
3.5. The Embedded Semantics
The size of the operator assigned prefix means that there is
potentially much more scope for embedding semantics than has
previously been possible. The following list describes some
suggested semantics which may be useful to network operators besides
source/destination location:
a. User types
b. Applications
c. Security domain
d. Traffic identity types
e. Quality requirements
f. Geo-location
The selection of semantics varies among different network operators.
They may choose one or more semantics to be embedded into their IPv6
address schemas, depending on what is important for them and what may
trigger packet differentiation processes in their networks. The
selection criterion and the impact of each choice are out of scope of
this document.
3.6. Network Operations Based on Semantic Prefixes
From the explicit semantics contained within the addresses of each
packet, many network operations can be applied. Compared with
traditional operations, these operations are easier to realize and
stable. Although detailed operation vary depending on various
embedded semantics, the network operations based on semantic prefix
can be abstracted into following categories:
a. Statistic based on certain semantic. Any embedded semantic can
be set as a statistic condition. In other words, any embedded
semantic can be measured independently.
b. Differentiate packet processing. Many packet processing
operations can be applied based on the semantic differentiation,
such as queueing, path selection, forwarding to certain process
devices, etc.
Jiang, et al. Expires January 16, 2014 [Page 8]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
c. Security isolation. A set of packet filters that are based on
semantic can fulfil network security isolation.
d. Access control. Resource access, authentication, service access
can be directly based on semantics.
e. Resource allocation. Resources, such as bandwidth, fast queue,
caching, etc., can be allocated or reserved for certain semantic
users/packets.
f. Virtualization. Within a Semantic Prefix Domain, organizing
virtual networks is simplified by assigning all the nodes the
same semantic identifier so that the packets from them can be
distinguished from other virtual networks.
It should also be noticed that these operations do not have to be
processed on the same single device. They may be separated among
network devices. In other words, if there are multiple semantics in
a Semantics Prefix Domain, various semantics may be understood and
treated on different network devices. It is not necessary for all
network devices in such domain to capable of understanding all
semantics.
4. Potential Benefits
Depending on various embedded semantics, different beneficial
scenarios can be expected.
a. Semantic prefix address schema provides a directly and explicitly
mechanism for packet inspection. It improves the inspecting
efficiency on IPv6 network devices.
b. Simplified measurement and statistics gathering: the semantic
prefix provides explicit identifiers which can be used for
measurement and statistical information collection. This can be
achieved by checking certain bits of the source and/or
destination address in each packet.
c. Simplified flow control: by applying policies according to
certain bit values, packets carrying the same semantics in their
source/destination addresses can.
d. Service segregation: when service related information is encoded
within the semantic prefix, this can be used to create simple
access-control lists which can be applied uniformly across all
network devices. Security zones are such typical services that
need to be segregated.
Jiang, et al. Expires January 16, 2014 [Page 9]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
e. Policy aggregation: the semantic prefix allows many policies to
be aggregated according to the same semantics within the policy
based routing system [RFC1104].
f. Easy dynamic reconfiguration of semantic oriented policy: network
operators may want to dynamically change the policy actions that
are operated on certain semantic packets. The semantic prefix
allows such changes be operated easily, as only a small number of
consistent policy rules need to be updated on all devices within
the semantic prefix domain.
g. Application-aware routing: embedding application information into
IP addresses is the simplest way to realize application aware
routing.
h. Easy user behavior management: based on the user type reading
from the addresses, any improper user behaviors can be easy
detected and automatically handle by network policies.
i. Easy network resources access rights management: the
authentication of access right may already be embedded into the
addresses. Simple matching policies can filter improper access
requests.
j. Easy virtualization: virtual network based on any semantics can
be easily deployed using the semantic prefix mechanism.
5. Potential Drawbacks
a. Address consumption caused by lower address utility rate.
Embedding semantics into IPv6 addresses causes the network to use
more of the address space that it normally would. The wastage
comes from aligning. 1) A small addressing requirement for a
separate type may get the same large address space as a large
addressing requirement. 2) The number of types in each semantic
has to align to 2^n, for example, 5 types uses to take 3 bits in
the prefix.
Network operators should be aware they may not get more addresses
because they have allocated their assigned address block(s) for
semantic use without the addresses actually being in use -
leading to a lower address utility rate. Although the current
Regional Internet Registry (RIR) policies do not disallow such
address usage, such usage has not been taken into account in
calculating reasonable addressing quotients.
b. Complexity that is created within the semantic prefix policy.
Encoding too many semantics into prefixes can come at the expense
Jiang, et al. Expires January 16, 2014 [Page 10]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
of future addressing flexibility. At the same time, embedding
too many semantics may induce semantic overlap. Careful
consideration should be taken with semantics definition.
c. The risk of privacy/information leakage. The semantics in the
address may be guessable, or leaked to outside the organisation.
Therefore, some information of either subscribers or networks may
be leaked, too.
d. Burdening the host OS. In some complex semantic prefix
scenarios, the semantics prefix mechanism puts extra burden on
the originator. In such scenarios, host devices are given
multiple IPv6 prefixes and required to choose correctly. When
forming a packet, the originator of packets (normally the host
OS) has to pick the right address/prefix according to the
semantics to access a service.
e. In order to perform policies based on trusted user/prefix, tight/
strict access control filter linked with prefix assignment is
requested. It is the filter who makes sure the prefix right.
The filter should link back to other states of the user, like
user authentication, etc, in order to match the packet to its
properties and check whether it is mapped to right semantics or
not.
6. Gaps for complex semantic prefix scenarios
The simplest semantic prefix model is to embed only abstracted user
type semantics into the prefix. Current network architectures can
support this semantic prefix model, in which each subscriber is still
assigned a single prefix, while they are not notified the semantic
embedded in the prefix.
In order to fulfill more benefits of the semantic prefix design,
additional functions are needed to allow semantic relevant operations
in networks and semantic relevant interactions with hosts.
IPv6 provides a facility for multiple addresses to be configured on a
single interface. This creates a precondition for the approach that
user chooses addresses differently for different purposes/usages.
6.1. Semantic Notification in the Network
In order to manage semantic prefixes and their relevant network
actions, the network should be able to notify semantics along with
prefix delegation.
Jiang, et al. Expires January 16, 2014 [Page 11]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
When an prefix is delegated using a DHCPv6 IA_PD [RFC3633], the
associated semantics should also be propagated to the requesting
router. This is particularly useful for autonomic process when a new
device is connected.
6.2. Semantic Relevant Interactions between Hosts and the Network
The more that semantics are embedded into a prefix, the more
complicated functions are needed for semantic relevant interactions
between hosts and the network, such as prefix delegation, host
notification, address selections, etc.
In practice, a single host may belong to multiple semantics. This
means that several IPv6 addresses are configured on a single physical
interface and should be selected for use depending on the service
that a host wishes to access. A certain packet would only serve a
certain semantic.
The host's IPv6 stack must have a mechanism for understanding these
semantics in order to select the right source address when forming a
packet. If the embedded semantic is application relevant,
applications on the hosts should also be involved in the address
choosing process: the host IPv6 stack reports multiple available
addresses to the application through socket API (one example is "IPv6
Socket API for Source Address Selection" [RFC5014]). The application
then needs to apply the semantic logic so that it can correctly
select from the offered candidate addresses.
Although [RFC6724] provides an algorithm for source address
selection, some semantic prefix policies may conflict with this
algorithm. In this case, source address selection mechanisms may
need further supporting functions to be developed.
6.3. Additional Technical Extensions
There are several areas in which the semantic prefix could be
extended in order to increase the usefulness and applicability of the
semantic prefix address schema. They are listed here for future
study. Currently, their feasibility, usefulness and applicability
are not carefully studied yet.
- Dynamic Policy Configuration
Dynamic policy configuration would simplify the distribution of
policy across devices in the semantic prefix domain. New functions
or protocol extension are needed to enable dynamic changes to the
policy actions in operation on certain semantic packets.
Jiang, et al. Expires January 16, 2014 [Page 12]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
- Semantics Announcements to peer networks
A network may announce all, or some of its Semantic Prefix Policy to
connected peer networks. This could be used to enable more dynamic
configuration and enable traffic from different semantic prefix
domains to traverse different networks whilst having the same
semantic prefix policy applied. To achieve this automatically by
message exchanging would require new functions or protocol
extensions.
- Extension of Prefix Semantics beyond the left-most 64 bits
The prefix concept refers here to the left-most bits in the IP
addresses delegated by the network management plane. The prefix
could be longer than 64-bits if the network operators strictly manage
the address assignment by using Dynamic Host Configuration Protocol
for IPv6 (DHCPv6) [RFC3315] (but in this case standard StateLess
Address AutoConfiguration - SLAAC [RFC4862] cannot be used).
- Organizing consumer/home networks according to semantics
Consumers or subscribers are currently assigned /48 or /56 prefixes.
They have bits, which may also count the right-most 64 bits too, to
organize their networks into subnets. These subnets may be organized
according to some semantics that are meaningful for the user himself.
In such scenario, the user acts as the network operator for his own
network. Some additional tecnologies/functions may be needed to make
such organizing and follow-up management efficient.
7. IANA Considerations
This document has no IANA considerations.
8. Change Log (removed by RFC editor)
draft-jiang-semantic-prefix-04: add new pitfalls section; restructure
to be a neatrul analysis document; 2013-07-15.
Jiang, et al. Expires January 16, 2014 [Page 13]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
draft-jiang-semantic-prefix-05: reword to emphasis this mechanism is
a (not the) method that network operators use their addresses; add
text to clarify the increased trust is actually from the deployment
of source address filter, which is a compliance requirement by
semantic prefix; restructure the document, move examples and gap
analysis into appendixes, reorganize most content into a frame
section; add summarized description for framework at the beginning of
Section 3; add description for network operations based on semantic
prefix; add a new coauthor who contributes an enterprise semantic
prefix network example; combine most of draft-sun-v6ops-semantic-
usecase into the draft as ISP example in appendix; 2013-5-28.
draft-jiang-semantic-prefix-04: add new coauthor, re-organize the
content, and refine the English, 2013-1-31.
draft-jiang-semantic-prefix-03: add the concept of hierarchical
Semantic Prefix Domain and more gap analysis, 2012-10-22.
draft-jiang-semantic-prefix-02: resubmitted to v6ops WG. Removed
detailed examples and recommendations for semantics bits, 2012-10-15.
draft-jiang-semantic-prefix-01: added enterprise considerations and
scenarios, emphasizing semantics only for local meaning and no intend
to standardize any common global semantics, 2012-07-16.
9. Security Considerations
Embedding semantics in prefix is actually exposing more information
of packets explicit. These informations may also provide convenient
for malicious attackers to track or attack certain type of packets.
If networks announce their local prefix semantics to their peer
networks, it may also increase the vulnerable risk.
Prefix-based filters should be deployed, in order to protect against
address spoofing attacks or denial of service for packets with forged
source addresses.
10. Acknowledgements
Useful comments were made by Erik Nygren, Dan Wing, Nick Hilliard,
Ray Hunter, David Farmer, Fred Baker, Joel Jaeggli, John Curran, Tim
Chown, Ted Lemon, Owen DeLong, Lorenzo Colitti, George Michaelson,
Joel Halpern, Vizdal Ales, Bless Roland, Manning Bill, Manfred Albert
and other participants in the V6OPS working group.
11. References
Jiang, et al. Expires January 16, 2014 [Page 14]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
11.1. Normative References
[RFC1104] Braun, H., "Models of policy based routing", RFC 1104,
June 1989.
[RFC2474] Nichols, K., Blake, S., Baker, F., and D. Black,
"Definition of the Differentiated Services Field (DS
Field) in the IPv4 and IPv6 Headers", RFC 2474, December
1998.
[RFC3315] Droms, R., Bound, J., Volz, B., Lemon, T., Perkins, C.,
and M. Carney, "Dynamic Host Configuration Protocol for
IPv6 (DHCPv6)", RFC 3315, July 2003.
[RFC3633] Troan, O. and R. Droms, "IPv6 Prefix Options for Dynamic
Host Configuration Protocol (DHCP) version 6", RFC 3633,
December 2003.
[RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing
Architecture", RFC 4291, February 2006.
[RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless
Address Autoconfiguration", RFC 4862, September 2007.
[RFC6724] Thaler, D., Draves, R., Matsumoto, A., and T. Chown,
"Default Address Selection for Internet Protocol Version 6
(IPv6)", RFC 6724, September 2012.
11.2. Informative References
[RFC5014] Nordmark, E., Chakrabarti, S., and J. Laganier, "IPv6
Socket API for Source Address Selection", RFC 5014,
September 2007.
Appendix A. An ISP Semantic Prefix Example
This ISP semantic prefix example is abstracted from a real ISP
address architecture design.
Note: for now, this example only covers unicast address within IP
Version 6 Addressing Architecture [RFC4291].
For ISPs, several motivations to use semantic prefixes are as
follows:
a. Network Device management: Separated and specialized address
space for network device will help to identify the network device
among numerous addresses and apply policy accordingly.
Jiang, et al. Expires January 16, 2014 [Page 15]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
b. Differentiated user management: In ISPs' network, different kinds
of customers may have different requirements for service
provisioning.
c. High-priority service guarantee: Different priorities may be
divided into apply differentiated policy.
d. Service-based Routing: ISPs may offer different routing policy
for specific service platforms .e.g.video streaming, VOIP, etc.
e. Security Control: For security requirement, operators need to
take control and identify of certain devices/customers in a quick
manner.
f. Easy measurement and statistic: The semantic prefix provides
explicit identifiers for measurement and statistic.
These requirements are largely falling into two categories: some is
regarding to the network device features, and the others are related
to services provision and subscriber identification. The functional
usage of the semantics for the two categories are quite different.
Therefore, an ISP semantic IPv6 prefix example is designed as a two-
level hierarchical architecture, in which the first level is the
function types of prefixes, and the second level is the further usage
within an specific prefix type.
A.1. Function Type Semantic Bits
Function Type (FT): the value of this field is to indicate the
functional usage of this prefix. The typical types for operators
include network device, subscriber and service platform.
+--------+--------+------------------------------------------------+
| | FT | |
+--------+--------+------------------------------------------------+
/ \
+------------+------------+
|000: network device |
|000~010: service platform|
|011~101: subscriber |
|110: reserved |
+-------------------------+
Function Type Bits Example
Figure 1
Jiang, et al. Expires January 16, 2014 [Page 16]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
The portion of each type should be estimated according to the actual
requirements for operators, in order to use the address space most
efficiently. Within the above FT design, the whole ISP IPv6 address
space is divided into four parts: the network device address space (1
/8 of total address space), the service platform address space (2/8
of total address space), the subscriber address space (3/8 of total
address space), and a reserved address space (1/8 of total address
space) for future usage.
A.2. Network Device Type Bits within Network Device Address Space
Network Device Type (NDT) indicates different types of network
devices. Normally, one operator may have multiple networks,
e.g.backbone network, mobile network, ISP brokered service network,
etc. Using NDT field to indicate specific network within an operator
may help to apply some routing policies. Locating NDT bits in the
left-most bits means that a single, simple access- control list
implemented across all networking devices would be enough to enforce
effective traffic segregation. The Locator field is followed behind
NDT.
+--------+--------+------+-----------------------------------------+
| | FT(000)| NDT | Locator | Network Device bits |
+--------+--------+------+-----------------------------------------+
/ \
/ \
+------------+-----+
|000~001: SubNet 1|
|010~110: SubNet 2|
|111: Reserved|
+------------------+
Network Device Type Bits Example
Figure 2
The portion of each subnet type should be estimated according to the
actual requirements for operators, in order to use the address space
most efficiently. Within the above NDT design, SubNet 1 is assigned
2/8 of the network device address space, SubNet 2 is assigned 5/8,
and 1/8 is reserved.
A.3. Subscriber Type Bits within Subscriber Address Space
Jiang, et al. Expires January 16, 2014 [Page 17]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
Subscriber Type (ST) indicates different types of subscribers, e.g.
wireline broadband subscriber, mobile subscriber, enterprise, WiFi,
etc. This type of prefix is allocated to end users. Further,
division may be taken on subscriber's priorities within a certain
subscriber type.
The Locator field within subscriber address space is put before ST
for better routing aggregation.
+--------+--------+---------------+------+-------------------------+
| | FT(011)| Locator | ST | Subscriber bits |
+--------+--------+---------------+------+-------------------------+
/ \
/ \
+----------+------------+---------------+
|0000~0011: broadband access subscriber |
|0100~0111: mobile subscriber |
|1000~1001: enterprise |
|1010~1011: WiFi subscriber |
|1100~1111: Reserved |
+---------------------------------------+
Subscriber Type Bits Example
Figure 3
The portion of each subscriber type should be estimated according to
the actual requirements for operators, in order to use the address
space most efficiently. Within the above ST design, the broadband
access subscriber type is assigned 4/16 of the subscriber address
space, the mobile subscriber is assigned 4/16, enterprise type and
WiFi subscriber type are assigned 2/16 each, and 2/16 is reserved.
A.4. Service Platform Type Bits within Service Platform Address Space
Service Platform Type (SPT) indicates typical service platforms
offered by operators. This field may have scalability problem since
there are numerous types of services . It is recommended that only
aggregated service platform types should be defined in this field.
This type of prefix is usually allocated to service platforms in
operator's data center.
+--------+--------+---------------+------+-------------------------+
| | FT(001)| Locator | SPT | Service bits |
+--------+--------+---------------+------+-------------------------+
/ \
/ \
+----------+------------+---------------+
Jiang, et al. Expires January 16, 2014 [Page 18]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
|000~001: Self-running service platform |
|001~011: Tenant service platform |
|100~101: Independent service platform |
|110~111: Reserved |
+---------------------------------------+
Service Platform Type Bits Example
Figure 4
The portion of each subnet type should be estimated according to the
actual requirements for operators, in order to use the address space
most efficiently.
Appendix B. An Enterprise Semantic Prefix example
This enterprise semantic prefix example is also abstracted from an
ongoing enterprise address architecture design. This example is
designed for a realtime video monitor network across a city region.
The semantic prefix solution is planning to be deployed along with a
strict authorization system.
Note: this example only covers unicast address within IP Version 6
Addressing Architecture [RFC4291].
For this example, the below semantics are important for the network
operation and require different network behaviors.
a. Terminal type: there are two terminal types only: monitor cameras
or video receivers. They are estimated to have similar number.
Network devices use another different address space.
b. Geographic location: the city has been managed in a three-level
hierarchical regionalism: district, area and street. Each level
has less than 28 sub-regions. This can also be considered as a
replacement of topology locator within this specific network.
c. Authorization level: the network operator is planning to
administrate the authorization in three or four levels. An
receiver can access the cameras that are the same or lower
authorization level.
d. Civilian or police/government.
e. Device attribute: this indicates the attribute of a camera
device. The attribute is expressed in an abstract way, such as
road traffic, hospital, nursery, bank, airport, etc. The
abstracted attribute type is designed to be less than 64.
Jiang, et al. Expires January 16, 2014 [Page 19]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
f. Receiver Attribute: this indicates the attribute of a video
receiver. The attribute is based on the receiver group, such as
police, firefighter, local security, etc. The attribute/receiver
group type is designed to be less than 128.
This example enterprise network has obtained a /32 address block from
ISP. There is another /48 dedicated for network devices.
The first bit is Terminal type, which indicates terminal type.
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ISP assigned block |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|T| Geographic Locator | AL|C|Device Attr| Device Bit |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
A semantic prefix design example for cameras
Figure 5
3-level hierarchical geographic locator takes 15 bits (each level 5
bits, 32 sub-regions). Authorization level takes 2 bits and 1 bit
differentiates civilian or police/government. 6 bits is assigned for
device attribute.
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ISP assigned block |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|T| GeoLoc | AL|C|Receiver Attr| Topology Locator |ReceiverBit|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
An semantic prefix design example for video receivers
Figure 6
The receiver is not as much as geographically distributed as cameras.
Therefore, Geographic locator is only detailed to district level.
Topology locator is needed for network forwarding and aggregation
within a district. It is assigned 10 bits. Authorization level bits
and civilian bit are the same with camera address space. Receive
attribute takes 7 bits, giving it is designed to be up to 128.
Appendix C. A Multi-Prefix Semantic example
A multiple-site enterprise may have been assigned several prefixes of
different lengths by its upstream ISPs. In this situation, in order
Jiang, et al. Expires January 16, 2014 [Page 20]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
to create a single, contiguous Semantic Prefix Domain, it is
necessary to base the semantic prefix policy on the longest assigned
prefix to ensure that there in enough addressing space to encode a
consistent set of semantics across all of the assigned prefixes.
In this example, an enterprise has received a /38 address block for
one site (A) and a /44 for a second site (B) . They can be organized
in the same Semantic Prefix Domain. The most-left 18 (site A) and 12
(site B) bits are allocated as locator. It provides topology based
network aggregation. The 8 right-most bits (from bits 56 to 63) are
assigned as the semantic field. In this design, the multiple-site
enterprise that has been assigned two prefixes of different lengths
can be organized as the same Semantic Prefix Domain. The semantic
and the Semantic Prefix Domain can traverse the intermediate ISP
networks, or even public networks.
The similar situation may happen on ISPs in the future, when an ISP
used up its assigned address space, or built up multiple networks in
different places.
Authors' Addresses
Sheng Jiang (editor)
Huawei Technologies Co., Ltd
Q14, Huawei Campus, No.156 BeiQing Road
Hai-Dian District, Beijing 100095
P.R. China
Email: jiangsheng@huawei.com
Qiong Sun
China Telecom
Room 708, No.118, Xizhimennei Street
Beijing 100084
P.R. China
Email: sunqiong@ctbri.com.cn
Ian Farrer
Deutsche Telekom AG
Bonn 53227
Germany
Email: ian.farrer@telekom.de
Jiang, et al. Expires January 16, 2014 [Page 21]
Internet-Draft Semantic IPv6 Prefix Analysis July 2013
Yang Bo
Huawei Technologies Co., Ltd
Q21, Huawei Campus, No.156 BeiQing Road
Hai-Dian District, Beijing 100095
P.R. China
Email: boyang.bo@huawei.com
Tianle Yang
China Mobile
32, Xuanwumenxi Ave. Xicheng District
Beijing 100053
China
Email: yangtianle@chinamobile.com
Jiang, et al. Expires January 16, 2014 [Page 22]
