A Simple Secure Addressing Generation Scheme for IPv6 AutoConfiguration (SSAS)
draft-rafiee-6man-ssas-03
The information below is for an old version of the document.
| Document | Type |
This is an older version of an Internet-Draft whose latest revision state is "Expired".
|
|
|---|---|---|---|
| Authors | Hosnieh Rafiee , Christoph Meinel | ||
| Last updated | 2013-03-11 | ||
| RFC stream | (None) | ||
| Formats | |||
| Stream | Stream state | (No stream defined) | |
| Consensus boilerplate | Unknown | ||
| RFC Editor Note | (None) | ||
| IESG | IESG state | I-D Exists | |
| Telechat date | (None) | ||
| Responsible AD | (None) | ||
| Send notices to | (None) |
draft-rafiee-6man-ssas-03
IPv6 maintenance Working Group (6man) H. Rafiee
INTERNET-DRAFT C. Meinel
Updates RFC 3971 , RFC 3972, RFC 4941 Hasso Plattner Institute
(if approved)
Intended status: Proposed Standard
Expires: September 11, 2013 March 11, 2013
A Simple Secure Addressing Generation Scheme for IPv6 AutoConfiguration
(SSAS)
<draft-rafiee-6man-ssas-03.txt>
Abstract
The default method for IPv6 address generation uses an
Organizationally Unique Identifier (OUI) assigned by the IEEE
Standards Association and an Extension Identifier assigned to the
hardware manufacturer [1] (section 2.5.1 RFC-4291) [RFC4291]. This
fact thus means that a node will always have the same Interface ID
(IID) whenever it connects to a new network. Because the node's IP
address does not change, the node will be vulnerable to privacy
related attacks. Currently this problem is addressed by the use of
two mechanisms that do not make use of the MAC address, or other
unique values that can be used for ID generation, for randomizing the
IID; Cryptographically Generated Addresses (CGA) [RFC3972] and
Privacy Extension [RFC4941]. The problem with the former approach is
the computational cost involved for the IID generation and in the
verification process. The problem with the latter approach is that it
lacks necessary security mechanisms and provides the node with only
partial protection against privacy related attacks. This document
proposes the use of a new algorithm for use in the generation of the
IID while, at the same time, securing the node against some types of
attack, like IP spoofing. These attacks are prevented by the addition
of a signature to messages sent over the network and by direct use of
a public key in the IP address.
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
Rafiee, et al. Expires September 11, 2013 [Page 1]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
This Internet-Draft will expire on September 11, 2013.
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. Conventions used in this document . . . . . . . . . . . . . . 3
3. Problem Statement . . . . . . . . . . . . . . . . . . . . . . 4
3.1. SSAS Applications . . . . . . . . . . . . . . . . . . . . 5
3.1.1. Preventing Attacks . . . . . . . . . . . . . . . . . 5
3.1.1.1. Replay attack . . . . . . . . . . . . . . . . . . 5
3.1.1.2. IP spoofing . . . . . . . . . . . . . . . . . . . 5
3.1.1.3. Denial of Service (DoS) attacks . . . . . . . . . 5
3.1.1.4. Spoofed Redirect Message . . . . . . . . . . . . 6
3.1.2. Nodes with limited resources . . . . . . . . . . . . 6
3.1.3. Other Applications . . . . . . . . . . . . . . . . . 7
4. Algorithms Overview . . . . . . . . . . . . . . . . . . . . . 7
4.1. Observe Privacy and Security . . . . . . . . . . . . . . 7
4.1.1. Interface ID (IID) Generation . . . . . . . . . . . . 7
4.1.2. Signature Generation . . . . . . . . . . . . . . . . 9
4.1.3. Generation of NDP Messages . . . . . . . . . . . . . 9
4.1.3.1. SSAS signature data field . . . . . . . . . . . . 10
4.1.4. SSAS verification process . . . . . . . . . . . . . . 11
4.2. Observe Privacy without Security Consideration . . . . . 13
4.2.1. Interface ID (IID) Generation . . . . . . . . . . . . 13
5. Security Considerations . . . . . . . . . . . . . . . . . . . 14
6. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 15
7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . 15
8. References . . . . . . . . . . . . . . . . . . . . . . . . . . 15
8.1. Normative . . . . . . . . . . . . . . . . . . . . . . . . 15
8.2. Informative . . . . . . . . . . . . . . . . . . . . . . . 16
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 17
Rafiee, et al. Expires September 11, 2013 [Page 2]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
1. Introduction
IPv6 addresses consist of two parts; the subnet prefix, which is the
64 leftmost bits of the IPv6 address, and the Interface ID (IID),
which is the 64 rightmost bits of the IPv6 address. The IEEE
Standards Association [1] (section 2.5.1 RFC-4291) [RFC4291] offered
a standard for the generation of IPv6 Interface IDs (IID) called the
Extended Unique Identifier (EUI-64). EUI-64s are generated by the
concatenation of an Organizationally Unique Identifier (OUI),
assigned by the IEEE Registration Authority (IEEE RA), with the
Extension Identifier assigned by the hardware manufacturer. For
example, if a manufacturer's OUI-36 hexadecimal value is
00-5A-D1-02-3, and the manufacture hexadecimal value, for the
Extension Identifier for a given component is 4-42-61-71, then the
EUI-64 value generated from these two numbers will be
00-5A-D1-02-34-42-61-71. If the OUI is 24 bits and the extension
identifier is also 24 bits (this constitutes the MAC address), then
to form the 64-bit EUI address, the OUI portion of the MAC address is
inserted into the leftmost 24 bits of the EUI-64 8 byte field and the
Extension Identifier is inserted into the rightmost 24 bits of the
EUI-64 8 byte field. A value of 0xFFFE is then inserted between these
two 24-bit items. IEEE has chosen 0xFFFE as a reserved value which
can only appear in an EUI-64 which is generated from an EUI-48 MAC
address. Bit 7 (u bit) in the OUI portion of the address is used to
indicate either global or local uniqueness. Globally unique addresses
assigned by the IEEE set this bit to zero, by default,indicating
global uniqueness.The bit is set to 1 for locally created addresses,
such as those used for virtual interfaces or a MAC address manually
configured by an administrator.
There are two mechanisms used to generate a randomized IID that do
not make use of a MAC address; CGA [RFC3972] and Privacy Extension
[RFC4941]. In this document we discuss the problem inherent with
using the current mechanisms and then we explain our solution to the
problem, which is to randomize the IID observing privacy, while, at
the same time, providing security to Neighbor Discovery Protocol
(NDP) messages of nodes in the IP layer. DHCPv6 [RFC3315] can also
benefit from this approach for the generation of a random IID or for
authentication purposes.
2. Conventions used in this document
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 [RFC2119].
In this document, these words will appear with that interpretation
only when in ALL CAPS. Lower case uses of these words are not to be
interpreted as carrying RFC-2119 significance.
Rafiee, et al. Expires September 11, 2013 [Page 3]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
In this document the use of || indicates the concatenation of the
values on either side of the sign.
3. Problem Statement
The drawback to using IIDs that do not change over time is one of
privacy. The node will generate the same IID whenever it joins a new
network thus making it easy for an attacker to track that node when
it moves to different networks.
The main problem with the privacy extension mechanism, when using the
first approach as explained in section 3.2.1 RFC-4941 [RFC4941],
i.e., using stable storage, is the lack of a provision for the use of
a security mechanism. The Privacy Extension RFC partially prevents
attacks related to privacy issues, but it cannot prevent attacks
related to security issues. For example, it cannot prevent IP
spoofing attacks and it cannot provide proof of IP address ownership
for a node. If one wants to use a secure method, with the privacy
extension, then one needs to use CGA. The problem with using CGA is
in the computational overhead necessary to compute it when higher sec
values are used and the time that is needed to perform the
verification process. This time is based on the reverse of the steps
required for the CGA regeneration during the verification process
along with the additional time needed for signature verification.
What is clear here is that it is not possible to generate the CGA
offline or before hand. This is because the subnet prefix (router
prefix) is one of the inputs to the SHA1 algorithm. The other problem
with CGA is the apparent lack of a defense against Denial of Service
(DoS) types of attack that are performed against verifier nodes. In
the CGA RFC there is no explanation as to how to prevent these types
of attacks. This means that an attacker can overwhelm the verifier
node with false CGA values thus rendering it unable to process
further messages.This document also proposes a solution to this type
of attack.
In order to overcome the problem with using the other mechanisms, the
time needed for IP address generation and verification needs to be
reduced. We propose the use of the SSAS algorithm, along with the
SSAS signature, to provide a node with the protection it needs to
protect it against IP spoofing and other spoofing types of attack in
the IP layer. Our experimental results [2] show that SSAS is much
faster than CGA when using a sec value of 0, and 600 times faster
than CGA when using a sec value of 1. This result will remain the
same in the future when there are faster CPUs because SSAS will also
benefit from the use of faster CPUs. Currently the generation time
for SSAS is less than 100 nanoseconds so when future new technologies
come into play it will be even less.
Note: It is not the intent of this document to obsolete CGA but to
Rafiee, et al. Expires September 11, 2013 [Page 4]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
propose a simpler and a faster addressing mechanism for use in
providing nodes with network layer privacy and security. This is
accomplished by providing a node with an algorithm to be used to
randomize the IID while at the same time providing nodes protection
against the types of attack explained below.
3.1. SSAS Applications
3.1.1. Preventing Attacks
The following sections detail some types of attack that SSAS can
prevent.
3.1.1.1. Replay attack
In this type of attack, an attacker will sniff the Neighbor Discovery
Protocol enabled network (NDP) messages to find, and then copy, a
legitimate signature and public key to his own NDP message which he
will then send to the original sender. But with the use of the SSAS
algorithm this can be prevented. With the addition of a timestamp and
by making a comparison of the public key to his own (the node's)
public key to that in the NDP message and also by the inclusion of
this timestamp in the signature. The use of a timestamp works because
the timestamp will be valid for only a short period of time. (this
accounts for clock skews.)
3.1.1.2. IP spoofing
This is a well-known type of attack in NDP. This type of attack is
used against the Duplicate Address Detection process. In this attack,
when a node joins the network and generates a new IP address, the
node sends a Neighbor Solicitation (NS) message to check for address
collisions in the network. The attacker, in this scenario, spoofs the
IP address and responds back to the node with a Neighbor
Advertisement (NA) message claiming ownership of this IP address.
While the SSAS algorithm does allow this node to verify other nodes
in the network, an attacker will not have the private key associated
with this node which is needed for SSAS signature generation, so the
verification process will fail.
3.1.1.3. Denial of Service (DoS) attacks
An attacker might send many NDP messages, using invalid signatures,
Rafiee, et al. Expires September 11, 2013 [Page 5]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
to a victim's node which then forces the node to busy itself with the
verification process. To mitigate this attack, a node SHOULD set a
limit on the number of messages (x) that should be verified within a
certain period of time. Implementations MUST provide a conservative
default and SHOULD provide a means for detecting when this limit is
reached.
3.1.1.4. Spoofed Redirect Message
Redirect messages, imitating the end host needing redirection, can be
sent from any router on the same broadcast segment. The attacker uses
the link-local address of the current first-hop router in order to
send a Redirect message to a legitimate node. Since that node
identifies the message as coming from its first hop router, by use of
the link-local address, it accepts the Redirect. The Redirect will
remain in effect as long as the attacker responds to the Neighbor
Unreachability Detection probes sent to the link-layer address. To
preclude this from occurring, the address ownership of the first-hop
router should be verified. The use of the SSAS verification process
will prevent such an attack.
3.1.2. Nodes with limited resources
SSAS can be used in nodes where limited computational resources are
available. It can provide protection to these nodes against the types
of attack stated above. Sensor networks are a prime example of nodes
with limited resources (such as battery, CPU, and etc); see RFC-4919
[RFC4919] for use in IPv6 networks. Because currently, as explained
in section 4. RFC-6775, the generation of the IID is based on EUI-64
which makes these nodes vulnerable to privacy and security attacks.
One of these types of attack can occur during the Duplicate Address
Detection (DAD) process.
Another example forthe use of SSAS would be in mobile networks during
the generation of IP addresses, as explained in section 4.4 RFC-6275
[RFC6275]. The current problem with the addressing mechanism in a
mobile node is that no privacy is observed when a node moves to
another network while usually keeping its Home Address. If there were
a fast and secure mechanism available, then it would be possible to
set this Home Address and change it and re-register it to the Home
network. Another possible use for SSAS in mobile nodes could be as a
security mechanism during the configuration of Care of Address (CoA);
see section 3. RFC-5213 [RFC5213]. In that RFC, home proxy plays the
role of a home agent for mobile nodes and mobile nodes set their CoA
by the use of either stateful or stateless autoconfiguration.
Currently they MUST use IPsec in order to secure this process.
Section 4 of that RFC discusses the possibility of using another
algorithm in order to secure mobile nodes.
Rafiee, et al. Expires September 11, 2013 [Page 6]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
3.1.3. Other Applications
With the wide usage of IP addresses in different types of devices and
by the use of autoconfiguration mechanisms to configure these IP
addresses, the need for the use of a security algorithm is increased.
One type of application would for use in vehicular networks or car by
car networks. There is currently some work in progress that makes use
of Neighbor Discovery. SSAS could also be a solution for enabling
fast protection against ND attacks.
4. Algorithms Overview
As explained earlier, one of the problems with using the current IID
generation approach is the compute intensive processing that is
needed for the IID algorithm generation. Another concern is for the
lack of security. Since we assume that a node will need to generate
and keep its address for a short period of time, we have tried to
keep the IID generation process to a minimum. We have also tried to
remain within the confines of NDP protocol. Here we offer two
algorithms. The first algorithm is used where both security and
privacy are a concern and the second algorithm is used where only
privacy is concern. The second algorithm addresses the problem with
the Privacy Extension RFC where the IID is not fully randomized. The
use of the first algorithm, where both network layer privacy and
security are observed, is RECOMMENDED.
4.1. Observe Privacy and Security
4.1.1. Interface ID (IID) Generation
To generate the IID a node will need to execute the following steps.
1. Generate a 2 byte (unsigned) random number called modifier. In the
generation of this modifier, implementations SHOULD use a random seed
in order to help in the randomization process of this number. Use of
a timestamp is OPTIONAL when a seed is used to generate the modifier.
The modifier MUST be greater than 128 (decimal). Set bits u and g
(bits 7 and 8) in the first byte of the modifier to zero.
2. Generate key pairs (public/private keys) using RSA or other
available algorithms. The default algorithm is RSA. Implementations
SHOULD store these keys in a safe place. It is RECOMMENDED that the
public key be generated, on the fly, during the start-up phase of the
algorithm generation. The keys MUST be valid over a certain period of
time. The implementation SHOULD consider this time based on the
Rafiee, et al. Expires September 11, 2013 [Page 7]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
network policy and privacy considerations. When the time expires for
the use of these key pairs, the node SHOULD generate new key pairs.
It SHOULD then use this new value for IP address and signature
generation. Implementation use of algorithms other that RSA is
OPTIONAL. Another algorithm that is RECOMMENDED for use in place of
RSA is ECC [3]. Comparing the use of ECC to that of RSA shows that
ECC with a 192 bit keys equivalent to a 1280 bit RSA key size. In
this case the packet size would be decreased by a factor 5 times
smaller than when RSA is used. However, with key sizes of 1024 bits
and 1280 bits, RSA generation and verification times are much faster
than when ECC is used. The other problem with the use of ECC is that
it might hold a patent and might not be royalty free.
3. Obtain the second byte of the modifier and call it the start field
pubkey. If the value of the start field pubkey is between 0 and the
size of public key length, in bytes, minus 6, then use this number as
an index for the public key array of bytes. Otherwise choose that
byte and shift its contents 2 bits to the right (the first two bits
will be zero) and set the start field pubkey to this number. This
ensures that the value of the start field pubkey will be less than
the size of the public key array of bytes, minus 6. This value helps
randomize the IID and minimizes the chance for a collision in the
network. For example, if the second byte of the modifier is 110, then
the start field pubkey value will be 110.
- If ECC is used for key generation the content of the start field
pubkey SHOULD be shifted 3 bits to the right. This insures that its
value will be less than the size of public key array of bytes, minus
6.
- If any future algorithms are used where the number of bytes in the
public key array of bytes is less than six bytes, then step 3 SHOULD
be skipped and the whole public key MUST used in concatenation with
the modifier for next step.
+-------------------------------------+
| | Pubkey | |
| | (48 bits) | |
+ +---------------+ +
| Public key |
| (1024 bits) |
+-------------------------------------+
Figure 1 Public key part of Interface ID
4. Concatenate the modifier with the six bytes from the public key
(starting at the start field pubkey) and call this the IID.
+---------------+----------------------+
| Modifier | Pubkey |
| (16 bits) | (48 bits) |
+---------------+----------------------+
Rafiee, et al. Expires September 11, 2013 [Page 8]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
Figure 2 Interface ID
5. Concatenate the IID with the local subnet prefix to set the local
IP address
6. Concatenate the IID with the router subnet prefix (Global subnet
prefix), obtained from the RA message, and set it as a tentative
global IP address. (This IP will become permanent after Duplicate
Address Detection (DAD) processing. (for more information about DAD
refer to section 4.1.3. )
4.1.2. Signature Generation
The SSAS signature is added to NDP messages in order to protect them
from IP spoofing and spoofing types of attack. SSAS will provide
proof of IP address ownership, as does the CGA generation algorithm,
but by using fewer steps. To generate the SSAS signature, the node
needs to execute the following steps:
1. Concatenate the timestamp with the 16 byte public key (that starts
at the start field pubkey) (see figure 3) and the global IP address.
The start field pubkey is one of the numbers that was introduced in
step 3 of section 4.1.1.
2. Sign the resulting value from step 1, using the RSA private key,
unless ECC or another available algorithm is used, and call the
resulting output the SSAS signature.
+---------+----------+-----------------+-------------+
|timestamp|Public key|Global IP Address|Other Options|
|(8 bytes)|(16 bytes)| (16 bytes) | (variable) |
+---------+----------+-----------------+-------------+
Figure 3 SSAS Signature
If NDP messages contain other data that must be protected, such as
important routing information, then this data SHOULD also be included
in the signature. The signature is designed for the inclusion of any
data needing protection. If there is no data that needs protection,
then the signature will only contain the timestamp, 16 byte public
key and Global IP address (Router subnet prefix plus IID).
4.1.3. Generation of NDP Messages
After a node generates its IP address, it should then process
Duplicate Address Detection in order to avoid address collisions in
the network. In order to do this the node needs to generate a
Rafiee, et al. Expires September 11, 2013 [Page 9]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
Neighbor Solicitation (NS) message. The format of a NS message is
shown in figure 4. The SSAS signature is added to the ICMPv6 options
of NS messages. The SSAS signature data field is an extended version
of the standard format of the RSA signature option of SEND [RFC3971].
The timestamp option is the same as that used with SEND. In the SSAS
signature, the data field contains the following items: type, length,
reserved, Other Len, pubkey len, public key, SSAS signature, and
padding.
+----------------+-------------+----------------------------+
| IPv6 Header |ICMPv6 header| ND message Specific Data |
| Next header= 58| | (variable) |
+--------------+-+-----------+-+----------------------------+
| Type = 13 | length | Reserved |
| (1 byte) | (1 byte) | (6 bytes) |
+--------------+-------------+------------------------------+
| timestamp |
| |
+----------+--------+---------+-----------+-----------------+
| Type = 12| length |Reserved | Other Len | algorithm type |
| (1 byte) |(1 byte)|(2 bytes)| (1 byte) | (1 byte) |
+----------+-----+--+--------------+------+-----------------+
| Subnet Prefix | Pubkey Len | Public Key in base64 |
| (8 byte) | (1 byte) | format |
+----------------+-----------------+------------------------+
| Other Options |
| |
+-----------------------------------------------------------+
| SSAS Signature |
| |
+-----------------------------------------------------------+
| padding |
| |
+-----------------------------------------------------------+
Figure 4 NDP Message Format with SSAS Signature Data Field
This document proposes an update to the SEND RFC in order to replace
the RSA signature field with the SSAS signature data field and to add
SSAS, as a new option, to SEND messages.
4.1.3.1. SSAS signature data field
- Type: This option should be set to 15. This number is a sequential
number in SEND options and represent the SSAS data field.
- Length: The length of the Signature Data field, including the Type,
Length, Reserved, pubkey Len, public key, Signature and padding,
Rafiee, et al. Expires September 11, 2013 [Page 10]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
should be a multiple of eight.
- Reserved: A 2 byte field reserved for future use. The value MUST be
initialized to zero by the sender, and MUST be ignored by the
receiver.
- Other Len: The length of other options in multiples of eight. The
length of this field is 1 byte.
- algorithm type: The algorithm used in generating key pairs and
signing the message. The length of this field is 1 byte. For RSA,
this value MUST be 0 and for ECC 1. For any other future algorithm it
MUST start from 2.
- Subnet Prefix: This is the router subnet prefix.
- PubKey Len. The length of the public key in multiples of eight.
- Public key. Base64 format of the public key
- Other Options. This variable-length field contains important data
that needs to be protected in the packet . The padding is used to
insure that the field is a multiple of eight in length.
- Padding. A variable-length field containing padding to insure that
the entire signature field is a multiple of eight in length. It thus
contains the number of blanks need to make the entire signature field
end on a multiple of eight.
All NDP messages should contain the SSAS signature data field which
allows receivers to verify senders. If a node receives a solicited NA
message in response to its NS message showing that another node
claims to own this address, then, after a successful verification
process, this node increments the modifier by one and again repeats
steps 2 thru 6 of section 4.1.1 . It repeats this process 3 times. If
after the third try the node receives the same claim, then it
considers it as an attack and will use that IP address.
4.1.4. SSAS verification process
A node's verification process should start when it receives NDP
messages.
Following are the steps used in the verification process:
1. Obtain the timestamp from the NDP message and call this value t1.
2. Obtain the timestamp from the node's system, convert it to UTC,
and call this value t2.
3. If (t2- x) < = t1 < = (t2 + x) go to step 4. Otherwise, the
Rafiee, et al. Expires September 11, 2013 [Page 11]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
message SHOULD be discarded without further processing. The value of
x is dependent on network delays and network policy. The
implementations MUST choose a flexible value for x based on the delay
in this network.
4. Obtain the public key from the SSAS signature data field.
5. Compare this to its own public key. If it is not the same, go to
the next step. Otherwise, the message should be discarded without
further processing. This step, along with the timestamp, can prevent
replay attacks.
6. Obtain the second byte of the modifier and call it the start field
pubkey. If the value of the start field pubkey is between 0 and the
size of public key length, in bytes, minus 6, then use this number as
an index for the public key array of bytes. Otherwise choose that
byte and shift its contents 2 bits to the right (the first two bits
will be zero) and consider this number the starting index of the
public key array of bytes. This ensures that the value of that byte
will be less than the size of the public key array of bytes, minus 6.
Set the start field pubkey to this number.
If ECC is used for key generation, then the content of the start
field pubkey SHOULD be shifted 3 bits to the right. This insures that
its value is less than the size of public key array of bytes, minus
6.
7. Obtain the IID from the sender's source IP address. (64 rightmost
bits of the IPv6 address)
8. Compare the 48 leftmost bits, starting at the start field pubkey
of the public key, to the 48 rightmost bits of the IID of the
sender's IP address. If they are the same, go to the next step.
Otherwise, the message should be discarded without further processing
9. Obtain the subnet prefix from the SSAS signature data field.
10. Concatenate the timestamp with the 16 bytes of the public key,
(starting from start field pubkey), the subnet prefix, the sender's
IID, and other options (if any) and call this entity the plain
message.
11. Obtain the SSAS signature from the SSAS signature data field.
Obtain the Algorithm type from the message.
12. Verify the Signature using the public key, and then enter the
plain message and the SSAS signature as an input to the verification
function. If the verification process is successful, process the
message. Otherwise, the message should be discarded without further
processing.
Rafiee, et al. Expires September 11, 2013 [Page 12]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
4.2. Observe Privacy without Security Consideration
4.2.1. Interface ID (IID) Generation
1. Generate a 16 byte random number called modifier. To generate this
modifier implementations SHOULD use a random seed to aid in the
randomization of this number.
2. Obtain the nodes' current time and convert it to timestamp. The
timestamp is a 64-bit unsigned integer field containing a timestamp.
The value indicates the number of seconds since January 1, 1970,
00:00 UTC, by using a fixed point format.
3. Concatenate the modifier with the timestamp.
R1=(modifier(16 bytes)||timestamp(8 bytes))
4. Execute SHA2 (256) on the result from step 3.
digest=SHA256(R1)
The use of SHA2 (256) is RECOMMENDED because the chances of finding a
collision are less than when using SHA1 and the generation time is
acceptable (in microseconds using a standard CPU). If, in the future,
a faster and collision free algorithm becomes available, then it
should be used. It is RECOMMENDED that the implementation be able to
support any new algorithms.
5. Generate a random number between 0 and 24 and call it the start
index. This number is used as an index for the SHA2 array of bytes.
This value helps randomize the IID and to minimize the chances for a
collision in the network. The length of this number is one byte.
If the output digest of future algorithms are less than 32, then the
random number must be the size of that digest, minus 8. This allows
the random number to be generated within the index of the digest
array of bytes.
6. Take the 64 leftmost bits (starting at the start index) from the
resulting output from step 5 (SHA2 digest) and set bits u and g (bits
7 and 8) to zero and call this the IID.
+-------------------------------------+
| | partial IID | |
| | (64 bits) | |
+ +---------------+ +
| SHA2 digest |
| (256 bits) |
+-------------------------------------+
Figure 5 Interface ID
Rafiee, et al. Expires September 11, 2013 [Page 13]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
7. Concatenate the IID with the local subnet prefix to set the local
IP address
8. Concatenate the IID with the router subnet prefix (Global subnet
prefix), obtained from the RA message, and set it as a tentative
global IP address. (This IP will become permanent after Duplicate
Address Detection (DAD) processing.
5. Security Considerations
As a security consideration what one might ask oneself is what are
the odds of an attacker being able to generate a public key having
four sequential bytes that are the same as the last rightmost 48 bits
of the IID; If he could, he could then generate the signature using
his own private key and thus break SSAS.
Mathematically it has been shown that the probability of matching 48
bits in the public key against 48 bits in the IID is about
pow(1/2,48) where pow is the power function, 2 is a base and 48 is an
exponent. Since the use of a public key and IP address with a maximum
lifetime of 30 days is RECOMMENDED, the probability of an attacker
finding the same value is 0.00000003, a very small value. When one
also considers the probability of an attacker being able to generate
a public key whose 48 bits, starting from an arbitrary point, matches
the 48 bits of the public key generated using the SSAS algorithm,
then the probability of his success is diminished even further. This
shows the strength of this algorithm against brute force attacks
while, at the same time, by using the signature and finding a binding
between the IP address and the public key, it provides proof of IP
address ownership at a speed that is about 600 times faster than that
of the CGA algorithm [2] (based on the implementation results, the
average time to generate SSAS including key generation is 250
microseconds).
Another consideration concerns routers wanting to use this algorithm
in place of CGA. As explained in RFC SEND, for routers, the use of a
Trusted Authority is RECOMMENDED along with the process of verifying
router certificates using other third party authorities (RFC-6494 and
RFC-6487) This will prevent a node from claiming to be a router. For
nodes then, rather than routers, SSAS can provide protection against
the types of attacks explained above.
Rafiee, et al. Expires September 11, 2013 [Page 14]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
6. IANA Considerations
This document defines a new algorithm for the generation of an
Interface ID in IPv6 networks that provides IP layer privacy and
local link security
7. Conclusions
Privacy has become a very important issue in recent years. A solution
for preventing a node from being tracked by an attacker is to change
the node's IP address frequently and by generating a random IID each
time a node wants to generate a new IP address. There are two
solutions available for randomizing the IID; CGA and Privacy
Extension. The former algorithm is compute intensive and the latter
algorithm is lacking in security. This document introduced a new
algorithm as a solution for providing privacy by randomizing the IID
and for providing security with the addition of a SSAS signature to
the NDP message and finding a binding between the public key and the
IP address. Our experimental results [2] show a definite improvement
in the computation time for the SSAS algorithm as compared to that
for the CGA algorithm. We also note that the probability of having
collisions with IP addresses, when using the SHA2 digest and the
public key, with a randomized 62 bit selection, approximates
pow(1/2,62) where pow is the power function, 2 is a base and 62 is a
exponent (u and g bits are ignored) . Moreover, the probability of an
attacker finding the public key which matches 48 rightmost bits of
the IID within 60 days approximates 0.00001 . This means this
algorithm is secure enough for wide usage.
8. References
8.1. Normative References
[RFC2119] Bradner, S., "Key words for use in RFCs to
Indicate Requirement Levels", BCP 14, RFC 2119, March 1997.
[RFC4291] Hinden, R., Deering, S., "IP Version 6 Addressing
Architecture," RFC 4291, February 2006.
[RFC3972] Aura, T., "Cryptographically Generated Addresses
(CGA)," RFC 3972, March 2005.
[RFC4941] Narten, T., Draves, R., Krishnan, S., "Privacy
Extensions for Stateless Address Autoconfiguration in
IPv6", RFC 4941, September 2007.
Rafiee, et al. Expires September 11, 2013 [Page 15]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
[RFC3971] Arkko, J., Kempf, J., Zill, B., and Nikander, P.,
"SEcure Neighbor Discovery (SEND)", RFC 3971, March 2005.
[RFC3315] Droms, R., Bound, J., Volz, B., Lemon, T.,
Perkins, C., Carney, M. , " Dynamic Host Configuration
Protocol for IPv6 (DHCPv6)", RFC 3315, July 2003.
[RFC4919] Kushalnagar, N., Montenegro, G., Schumacher, C.,"
IPv6 over Low-Power Wireless Personal Area Networks
(6LoWPANs): Overview, Assumptions, Problem Statement, and
Goals", RFC 4919, August 2007.
[RFC6775] Shelby, Z., Chakrabarti, S., Nordmark, E.,
Bormann, C. , " Neighbor Discovery Optimization for IPv6
over Low-Power Wireless Personal Area Networks (6LoWPANs)",
RFC 6775, November 2012.
[RFC6275] Perkins, C., Johnson, D., Arkko, J., "Mobility
Support in IPv6", RFC 6275, July 2011.
[RFC6543] Gundavell, S., "Reserved IPv6 Interface
Identifier for Proxy Mobile IPv6", RFC 6543, May 2012.
8.2. Informative References
[1] IEEE Standards Association,
http://standards.ieee.org/develop/regauth/tut/eui64.pdf, 2012
[2] Rafiee, H., "Research Results",
http://ipv6sra.rozanak.com/Jan2013_CGA_SSAS_Comparison.pdf, 2013
[3] Brown, R., L., D. : SEC 1: Elliptic Curve Cryptography,
Certicom Research,
http://www.secg.org/download/aid-780/sec1-v2.pdf, 2009
Rafiee, et al. Expires September 11, 2013 [Page 16]
INTERNET DRAFT SSAS for Autoconfiguration March 11, 2013
Authors' Addresses
Hosnieh Rafiee
Hasso-Plattner-Institute
Prof.-Dr.-Helmert-Str. 2-3
Potsdam, Germany
Phone: +49 (0)331-5509-546
Email: ietf@rozanak.com
Dr. Christoph Meinel
(Professor)
Hasso-Plattner-Institute
Prof.-Dr.-Helmert-Str. 2-3
Potsdam, Germany
Email: meinel@hpi.uni-potsdam.de
Rafiee, et al. Expires September 11, 2013 [Page 17]