Network Working Group J. Arkko Internet-Draft Ericsson Expires: August 25, 2003 J. Kempf DoCoMo Communications Labs USA B. Sommerfeld SUN Microsystems B. Zill Microsoft February 24, 2003 SEcure Neighbor Discovery (SEND) draft-ietf-send-ipsec-00.txt Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet-Drafts. 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." The list of current Internet-Drafts can be accessed at http:// www.ietf.org/ietf/1id-abstracts.txt. The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. This Internet-Draft will expire on August 25, 2003. Copyright Notice Copyright (C) The Internet Society (2003). All Rights Reserved. Abstract IPv6 nodes use the Neighbor Discovery (ND) protocol to discover other nodes on the link, to determine each other's link-layer addresses, to find routers and to maintain reachability information about the paths to active neighbors. If not secured, ND protocol is vulnerable to various attacks. This document specifies an extension to IPsec for securing ND. Contrary to the original ND specifications that also Arkko, et al. Expires August 25, 2003 [Page 1] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 called for use of IPsec, this extension does not require the creation of a large number of manually configured security associations. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4 2. Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3. Neighbor and Router Discovery Overview . . . . . . . . . . . 6 4. Secure Neighbor Discovery Overview . . . . . . . . . . . . . 9 5. Cryptographically Generated Addresses . . . . . . . . . . . 11 5.1 Address Format . . . . . . . . . . . . . . . . . . . . 12 5.2 Basic Interface Identifier Generation . . . . . . . . 12 5.3 Address Generation . . . . . . . . . . . . . . . . . . 13 5.4 Duplicate Address Detection . . . . . . . . . . . . . 14 6. Authorization Delegation Discovery . . . . . . . . . . . . . 15 6.1 Delegation Chain Solicitation Message Format . . . . . 15 6.2 Delegation Chain Advertisement Message Format . . . . 17 6.3 Trusted Root Option . . . . . . . . . . . . . . . . . 19 6.4 Certificate Option . . . . . . . . . . . . . . . . . . 20 6.5 Processing Rules for Routers . . . . . . . . . . . . . 21 6.6 Processing Rules for Hosts . . . . . . . . . . . . . . 22 7. IPsec Extensions . . . . . . . . . . . . . . . . . . . . . . 25 7.1 The AH_RSA_Sig Transform . . . . . . . . . . . . . . . 25 7.1.1 Reserved SPI Number . . . . . . . . . . . . . . 25 7.1.2 Authentication Data Format . . . . . . . . . . . 25 7.1.3 AH_RSA_Sig Security Associations . . . . . . . . 27 7.1.4 Replay Protection . . . . . . . . . . . . . . . 28 7.1.5 Processing Rules for Senders . . . . . . . . . . 28 7.1.6 Processing Rules for Receivers . . . . . . . . . 29 7.2 Other IPsec Extensions . . . . . . . . . . . . . . . . 30 7.2.1 Destination Agnostic Security Associations . . . 30 7.2.2 ICMP Type Specific Selectors . . . . . . . . . . 31 8. Securing Neighbor Discovery with SEND . . . . . . . . . . . 32 8.1 Using IPsec to Secure Neighbor Advertisement Messages 32 8.2 Security Policy and SA Database Configuration . . . . 32 9. Securing Router Discovery with SEND . . . . . . . . . . . . 34 9.1 Using IPsec to Secure Router Advertisement Messages . 34 9.2 Using IPsec to Secure Redirect Messages . . . . . . . 34 9.3 Security Policy and SA Database Configuration . . . . 35 10. Operational Considerations . . . . . . . . . . . . . . . . . 37 11. Performance Considerations . . . . . . . . . . . . . . . . . 39 12. Security Considerations . . . . . . . . . . . . . . . . . . 40 12.1 Achieved Security Properties . . . . . . . . . . . . . 40 12.2 Attacks against SEND Itself . . . . . . . . . . . . . 40 13. IANA Considerations . . . . . . . . . . . . . . . . . . . . 42 14. Conclusions and Remaining Work . . . . . . . . . . . . . . . 43 Normative References . . . . . . . . . . . . . . . . . . . . 44 Informative References . . . . . . . . . . . . . . . . . . . 45 Arkko, et al. Expires August 25, 2003 [Page 2] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . 46 A. Contributors . . . . . . . . . . . . . . . . . . . . . . . . 47 B. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 48 C. IPR Considerations . . . . . . . . . . . . . . . . . . . . . 49 Intellectual Property and Copyright Statements . . . . . . . 50 Arkko, et al. Expires August 25, 2003 [Page 3] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 1. Introduction IPv6 defines the Neighbor Discovery (ND) protocol in RFC 2461 [6]. Nodes on the same link use the ND protocol to discover each other's presence, to determine each other's link-layer addresses, to find routers and to maintain reachability information about the paths to active neighbors. The ND protocol is used both by hosts and routers. Its functions include Router Discovery (RD), Address Auto- configuration, Address Resolution, Neighbor Unreachability Detection (NUD), Duplicate Address Detection (DAD), and Redirection. RFC 2461 called for the use of IPsec for protecting the ND messages. However, it turns out that in this particular application IPsec can only be used with a manual configuration of security associations due to chicken-and-egg problems [17] in using IKE [15] before ND is operational. Furthermore, the number of security associations needed for protecting ND is impractically large [18]. Finally, RFC 2461 did not specify detailed instructions for using IPsec to secure ND. Section 4 describes our overall approach to securing ND. This approach involves the use of IPsec AH [3] to secure all advertisements relating to Neighbor and Router Discovery. A new transform for AH allows public keys to be used. Routers are certified by a trusted root, and a zero-configuration mechanism for showing address ownership. Section 5 describes the mechanism for showing address ownership, based on the use of Cryptographically Generated Addresses (CGAs). Section 6 describes the mechanism for distributing certificate chains to establish authorization delegation chain to a common trusted root. Section 7 describes the necessary modifications to IPsec. Finally, Section 8 show how to apply these components to securing Neighbor and Router Discovery. Arkko, et al. Expires August 25, 2003 [Page 4] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 2. Terms Cryptographically Generated Addresses (CGAs) A technique [22] where the address of the node is cryptographically generated from the public key of the node and some other parameters using a one-way hash function. Internet Control Message Protocol version 6 (ICMPv6) The IPv6 control signaling protocol. Neighbor Discovery is a part of ICMPv6. Neighbor Discovery (ND) The IPv6 Neighbor Discovery protocol [6]. Security Association (SA) A Security Association (SA) is a simplex "connection" that affords security services to the traffic carried by it. Security services are afforded to an SA by the use of AH, or ESP, but not both. A SA is uniquely identified by a triple consisting of a Security Parameter Index (SPI), an IP Destination Address, and a security protocol (AH or ESP) identifier [2]. Security Association Database (SAD) A nominal database containing parameters that are associated with each (active) security association. For inbound and outbound IPsec processing, these databases are separate. Security Parameters Index (SPI) An arbitrary 32-bit value. Together with the destination IP address and security protocol (ESP or AH) identifier, the SPI uniquely identifies the Security Association. Values from 1 to 255 are reserved. Special SPI A Security Parameters Index from the reserved range 1 to 255. Security Policy The Security Policy determines the security services afforded to an IPsec protected packet and the treatment of the packet in the network. Security Policy Database (SPD) A nominal database containing a list of policy entries. Each policy entry is keyed by one or more selectors that define the set of IP traffic encompassed by this policy entry. Separate entries for inbound and outbound traffic is required [2]. Arkko, et al. Expires August 25, 2003 [Page 5] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 3. Neighbor and Router Discovery Overview IPv6 Neighbor and Router Discovery have several functions. Many of these functions are overloaded on a few central message types such as the ICMPv6 Neighbour Discovery message. In this section we explain some of these tasks and their effects in order to understand better how the messages should be treated. Where this section and the original Neighbor Discovery RFCs are in conflict, the original RFCs take precedence. In IPv6, many of the tasks traditionally done at lower layers such as ARP have been moved to the IP layer. As a consequence, unified mechanisms can be applied across link layers, security mechanisms or other extensions can be adopted more easily, and a clear separation of the roles between the IP and link layer can be achieved. The main functions of IPv6 Neighbor Discovery are as follows: o Neighbor Unreachability Detection (NUD) is used for tracking the reachability of neighbors, both hosts and routers. NUD is defined in Section 7.3 of RFC 2461 [6]. No higher level traffic can proceed if this procedure flushes out neighbour cache entries after (perhaps incorrectly) determining that the peer is not reachable. o Duplicate Address Detection (DAD) is used for preventing address collisions [7]. A node that intends to assign a new address to one of its interfaces runs first the DAD procedure to verify that other nodes are not using the same address. Since the outlined rules forbid the use of an address until it has been found unique, no higher layer traffic is possible until this procedure has completed. o Address Resolution is similar to IPv4 ARP [14]. The Address Resolution function resolves a node's IPv6 address to the corresponding link-layer address for nodes on the link. Address Resolution is defined in Section 7.2 of RFC 2461 [6] and it is used for hosts and routers alike. Again, no higher level traffic can proceed until the sender knows the hardware address of the destination node or the next hop router. Note that like its predecessor in ARP, IPv6 Neighbor Discovery does not check the source link layer address against the information learned through Address Resolution. This allows for an easier addition of network elements such as bridges and proxies, and eases the stack implementation requirements as less information needs to be passed from layer to another layer. o Address Autoconfiguration is used for automatically assigning Arkko, et al. Expires August 25, 2003 [Page 6] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 addresses to a host [7]. This allows hosts to operate without configuration related to IP connectivity. The Address Autoconfiguration mechanism is stateless, where the hosts use prefix information delivered to them during Router Discovery to create addresses, and then test these addresses for uniqueness using the DAD procedure. A stateful mechanism, DHCPv6 [19], provides additional Autoconfiguration features. Router and Prefix Discovery and Duplicate Address Detection have an effect to the Address Autoconfiguration tasks. o The Redirect function is used for automatically redirecting hosts to an alternate router. Redirect is specified in Section 8 of RFC 2461 [6]. It is similar to the ICMPv4 Redirect message [13]. o The Router Discovery function allows IPv6 hosts to discover the local routers on an attached link. Router Discovery is described in Section 6 of RFC 2461 [6]. The main purpose of Router Discovery is to find neighboring routers that are willing to forward packets on the behalf of hosts. Prefix discovery involves determining which destinations are directly on a link; this information is necessary in order to know whether a packet should be sent to a router or to the destination node directly. Typically, address autoconfiguration and other tasks can't proceed until suitable routers and prefixes have been found. The Neighbor Discovery messages follow the ICMPv6 message format and ICMPv6 types from 133 to 137. The IPv6 Next Header value for ICMPv6 is 58. The actual Neighbor Discovery message includes an ND message header consisting of ICMPv6 header and ND message-specific data, and zero or more ND options: <------------ND Message-----------------> *-------------------------------------------------------------* | IPv6 Header | ICMPv6 | ND message- | ND Message | | Next Header = 58 | Header | specific | Options | | (ICMPv6) | | data | | *-------------------------------------------------------------* <--ND Message header---> The ND message options are formatted in the Type-Length-Value format. All IPv6 ND protocol functions are realized using the following messages: Arkko, et al. Expires August 25, 2003 [Page 7] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 ICMPv6 Type Message ------------------------------------ 133 Router Solicitation (RS) 134 Router Advertisement (RA) 135 Neighbor Solicitation (NS) 136 Neighbor Advertisement (NA) 137 Redirect The functions of the ND protocol are realized using these messages as follows: o Router Discovery uses the RS and RA messages. o Duplicate Address Detection uses the NS and NA messages. o Address Autoconfiguration uses the NS, NA, RS, and RA messages. o Address Resolution uses the NS and NA messages. o Neighbor Unreachability Detection uses the NS and NA messages. o Redirect uses the Redirect message. The addresses used in these messages are as follows: o Neighbor Solicitation: The destination address is either the solicited node multicast address, unicast address, or an anycast address. o Neighbour Advertisement: The destination address is either a unicast address or the All Nodes multicast address [1]. o Router Solicitation: The destination address is typically the All Routers multicast address [1]. o Router Advertisement: The destination address can be either a unicast or the All Nodes multicast address [1]. Like the solicitation message, the advertisement is also local to the link only. o Redirect: This message is always sent from the router's link-local address to the source address of the packet that triggered the Redirect. Hosts verify that the IP source address of the Redirect is the same as the current first-hop router for the specified ICMP Destination Address. Rules in [1] dictate that unspecified, anycast, or multicast addresses may not be used as source addresses. Therefore, the destination address will always be a unicast address. Arkko, et al. Expires August 25, 2003 [Page 8] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 4. Secure Neighbor Discovery Overview IPsec AH is used in to protect Neighbor and Router Discovery messages. This specification introduces the use of a new transform for IPsec AH, extensions to the current IPsec selectors, an authorization delegation discovery process, and an address ownership proof mechanism. The components of the solution specified in this document are as follows: o IPsec AH is used to protect all advertisement messages relating to Neighbor and Router discovery. Solicitation messages are not protected, as they do not carry any information. o IPsec security policy database and security association database are configured to require the protection as indicated above. Note that such configuration may take place manually or the operating system may perform it automatically upon enabling Secure Neighbor Discovery. This specification introduces extensions to the required selectors used in security policy database entries. This is necessary in order to enable the protection of specific ICMP message types, while leaving other messages unprotected. o A new transform for IPsec AH allows public keys to be used on a security association directly without the involvement of a key management protocol. Symmetric session keys are not used, public key signatures are used instead. The trust to the public key is established either with the authorization delegation process or the address ownership proof mechanism, depending on configuration and the type of the message protected. The new transform uses also a fixed, standardized SPI (Security Parameters Index) number. This necessary again in order to avoid the involvement of a key management protocol. Given that Neighbor and Router Discovery messages are in some cases sent to multicast addresses, the new transform uses a timestamp mechanism as a replay mechanism instead of sequence numbers. o Trusted roots are expected to certify the authority of routers. A host and a router must have at least one common trusted root before the host can accept adopt the router as its default router. Delegation Chain Solicitation and Advertisement messages are used to discover a certificate chain to the trusted root without Arkko, et al. Expires August 25, 2003 [Page 9] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 requiring the actual Router Discovery messages to carry lengthy certificate chains. o Cryptographically Generated Addresses are used to assure that the sender of a Neighbor or Router Advertisement is the owner of an the claimed address. A public-private key pair needs to be generated by all nodes before they can claim an address. Arkko, et al. Expires August 25, 2003 [Page 10] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 5. Cryptographically Generated Addresses Cryptographically Generated Addresses (CGAs) [22][23][20][25] are a technique whereby a node's IPv6 address can be unalterably tied to the node's public key. Conceptually, CGAs allow a recipient of a message to determine whether the sender is authorized to use the public key and address claimed to be associated with the packet. Typically, this requires the sender to use the hash of the node's public key as the interface identifier in the bottom 64 bits of the IPv6 address. Authorization through CGAs and certificates are related, but separate mechanisms. It is separate in that other techniques of authorization (i.e. digital certificates) can be used instead of CGAs to achieve the same effect. However, certificates require a means to create and distribute them, thereby imposing more overhead than CGA. It is related in that a digital signature is required in addition to the CGA address and the signature must cover the address, in order that the recipient can have the confidence that the address was not altered in transit. Furthermore, to properly authorize the address use, the issuer of the certificate must be considered as a valid source of authority for certifying address usage, and must be capable of making statements about an individual's use of IP addresses. Theoretically, proper use of certificates provides more assurance about address usage authorization than CGA. However, it is often practically difficult to arrange the certificate authorities so that they can control which IP addresses can be used by which parties. The authorization provide by CGA is computational in nature, deriving its strength from the computational difficulty of creating duplicate CGA addresses. It does not require any configuration tasks, and it does not impose any requirements on the infrastructure. Respectively, certificate based authorization is administrative in nature, and does not impose restrictions to the structure of the addresses. CGAs are particularly useful for Neighbor Discovery because they provide a low overhead way for the sender of a Neighbor Advertisement to indicate their authorization for claiming the address. The recipient of a Neighbor Advertisement with a CGA Address, a public key, and a digital signature in the header can have confidence that: o The packet was not modified in transit (due to the signature), o The sender of the packet has a right to claim possession of the address (due to the authenticated CGA address). In this section, we describe how a sender generates CGA addresses and digital signatures for the AH header in Neighbor Advertisement Arkko, et al. Expires August 25, 2003 [Page 11] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 packets, and how the receiver of such a packet verifies it. The description of CGA use for IPv6 Neighbor Discovery follows closely that described in [24]. 5.1 Address Format The basic idea behind CGA addresses is to use some function of the host's public key as input to a hash function to generate the interface identifier (bottom 64 bits) in the IPv6 address. Variations on this basic theme provide additional security against denial of service attacks and futureproofing against increases in attacker processing power due to Moore's Law. For purposes of secure Neighbor Discovery, CGA addresses are modified EUI-64 addresses [1] in which the "universal/local" bit (bit 6) is set to 1 (indicating global scope) and the "individual/group" bit (bit 7) is set to 1 (indicating the CGA group). Correct handling of these bits effectively reduces the size of the interface identifier to 62 bits. 5.2 Basic Interface Identifier Generation The basic hash algorithm for CGA addresses generates a 160 bit hash by concatenating the node's public key, a nonce, and routing prefix for the address in question. This result is then hashed to obtain the actual interface identifier. The input hash is generated as follows: Equation (1). H(N) = Hash-160(public_key | nonce | routing_prefix) H(i) = Hash-160(H(i+1)) where Hash-160 is the 160 bits obtained from applying the SHA-1 secure hashing algorithm [12], public_key is the node's public key in the format defined in Section 7.1.2, and nonce is a random octet string of 8 or more bytes. The selection N is a local matter, but it MUST be at least 3. N is used as a defense mechanism against denial-of-service attacks. The routing_prefix is the routing prefix for the address in question. Note that this value is used regardless of whether the scope of the address to be generated is link-local, site-local, or global. If the scope is not global, it is possible that different networks will be using the same routing prefix, such as the FE80::/10 prefix for link-local addresses. This is allowed, as the addresses are not used in the same network. In any case, other components in Equation (1) typically provide sufficient randomness to avoid collisions and Duplicate Address Detection would avoid possibly remaining address Arkko, et al. Expires August 25, 2003 [Page 12] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 collisions. The host generates a series of these hash values. The actual interface identifier is then generated by performing taking the rightmost 60 bits of the SHA-1 hash applied to the input value: Equation (2). interface_id = Hash-60(H(i)) where Hash-60 is the rightmost 60 bits obtained from the application of the SHA-1 algorithm, i starts at 0 and increases depending on whether additional rounds of duplicate address detection must be negotiated (see Section 5.4). The routing_prefix term is included in the above in order to introduce a strong binding between the prefixes and interface identifiers, and to add some randomness in order to defeat brute force and birthday attacks. If it is not included, an attacker can generate a lookup table of key pairs for each of the possible 2**60 values of the interface identifier and use them to disrupt duplicate address detection. Note, however, that including these in the address requires the host to perform duplicate address detection for each address configured on the interface, not just for the link local address, as is allowed by RFC 2462. 5.3 Address Generation Equation (2) in Section 4.1.1 only takes 60 bits of the SHA-1 hash, although 62 bits are theoretically available after the "u" and "g" bits are omitted. This is because the right most 2 bits of the interface identifier are reserved for a security parameter. The security parameter can have a value of 0 through 3, and is a way of future-proofing the CGA address against increases in processing power in attackers due to Moore's Law, since 62 bits is on the borderline of what is today computationally difficult to attack. Conceptually, the security parameter is a way to increase the computational effort of both generating and attacking an address. While this has the side effect of increasing the effort for the client, the client presumably only has to generate the address once, while an attacker may have to generate the address multiple times. The algorithm for generating the actual address is as follows, given the security parameter has value Sec: 1. Generate a key pair. 2. Generate a nonce value. Arkko, et al. Expires August 25, 2003 [Page 13] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 3. Generate a table of hash values according to Equation (1). 4. Generate a target identifier according to Equation (2), but taking 20 x Sec + 60 bits instead of just 60 bits. 5. Compare the leftmost 20 x Sec to zero. If not zero, go back to Step 2. 6. Set the universal and group bits to 1 and the rightmost two bits to Sec. 7. Use the result of Step 6 as the interface identifier for the address. If the security parameter is zero, this algorithm is the basic CGA algorithm. If the security parameter is greater than zero, the algorithm is not guaranteed to terminate after a certain number of iterations (though it will ultimately terminate). For security parameter values 1, 2, and 3, the average number of iterations required to produce a matching hash output are 2**19, 2**39, and 2**59, i.e. 2**(20 x Sec -1) [24]. The additional amount of computational effort involved in increasing the security parameter allows the SEND algorithm to scale as Moore's law increases processing power. 5.4 Duplicate Address Detection During Duplicate Address Detection, a node may encounter a clash with another node on the link. One possible denial of service attack occurs when the attacker deliberately provokes an address clash, in order to prevent the victim from claiming the address. RFC 2462 [7] inadvertently facilitates this attack, by requiring nodes to terminate Duplicate Address Detection when a clash is detected. For Secure Neighbor Discovery, a node performs Duplicate Address Detection a maximum of 3 times. If an address clash is detected, the node restarts interface identifier generation at Step 2 of the algorithm described in Section 4.1.2, by selecting a different hash input for target identifier generation. If clashes are detected after three tries, the node is probably under attack, so it should shut down and report the situation to an administrator. Arkko, et al. Expires August 25, 2003 [Page 14] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 6. Authorization Delegation Discovery Several protocols, including IPv6 Neighbor Discovery, allow a node to automatically configure itself based on information it learns shortly after connecting to a new link. It is particularly easy for "rogue" routers to be configured, and it is particularly difficult for a network node to distinguish between valid and invalid sources of information when the node needs this information before to communicate off-link. Since the newly-connected node likely can't communicate off-link, it can't be responsible for searching information to help validate the router; however, given a chain of appropriately signed certificates, it can check someone else's search results and conclude that a particular message comes from an authorized source. Similarly, the router, which is already connected to the network, can if necessary communicate off-link and construct the certificate chain. The Secure Neighbor Discovery protocol introduces two new ICMPv6 messages that can be used between hosts and routers to allow the client to learn the certificate chain with the assistance of the router. Where hosts have certificates from a trusted root, these messages may also optionally be used between hosts to acquire the peer's certificate chain. The Delegation Chain Solicitation message is sent by hosts when they wish to request the certificate chain between a router and the one of the hosts' trusted roots. The Delegation Chain Advertisement message is sent as an answer to this message, or periodically to the All Nodes multicast address. Due to the size of certificates and potentially long certificate chains, the advertisement message may be large. The messages have been made separate from the rest of Neighbor Discovery in order to reduce their effect on the size of other messages. Long certificate chains may also be broken to multiple messages. The Authorization Delegation Discovery process does not exclude other forms of discovering the certificate chains. For instance, during fast movements mobile nodes may learn information - including the chains - of the next router from the previous router. 6.1 Delegation Chain Solicitation Message Format Hosts send Delegation Chain Solicitations in order to prompt routers to generate Delegation Chain Advertisements quickly. Arkko, et al. Expires August 25, 2003 [Page 15] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 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 | Code | Checksum | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Identifier | Reserved | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Options ... +-+-+-+-+-+-+-+-+-+-+-+- IP Fields: Source Address An IP address assigned to the sending interface, or the unspecified address if no address is assigned to the sending interface. Destination Address Typically the all-routers multicast address, or the address of the hosts' default router. Hop Limit 255 ICMP Fields: Type TBD for Delegation Chain Solicitation. Code 0 Checksum The ICMP checksum [8].. Identifier This 16 bit unsigned integer field acts as an identifier to help match advertisements to solicitations. The Identifier field MUST NOT be zero. Arkko, et al. Expires August 25, 2003 [Page 16] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Reserved This field is unused. It MUST be initialized to zero by the sender and MUST be ignored by the receiver. Valid Options: Trusted Root One or more trusted roots that the client is willing to accept. Future versions of this protocol may define new option types. Receivers MUST silently ignore any options they do not recognize and continue processing the message. 6.2 Delegation Chain Advertisement Message Format Routers send out Delegation Chain Advertisement messages periodically, or in response to a Delegation Chain Solicitation. 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 | Code | Checksum | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Identifier |M| Component | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Reserved | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Options ... +-+-+-+-+-+-+-+-+-+-+-+- IP Fields: Source Address MUST be the link-local address assigned to the interface from which this message is sent. Destination Address Typically the Source Address of a host invoking Delegation Chain Solicitation or the all-nodes multicast address. Hop Limit 255 Arkko, et al. Expires August 25, 2003 [Page 17] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 ICMP Fields: Type TBD for Delegation Chain Advertisement. Code 0 Checksum The ICMP checksum [8].. Identifier This 16 bit unsigned integer field acts as an identifier to help match advertisements to solicitations. The Identifier field MUST be zero for unsolicited advertisements and MUST NOT be zero for solicited advertisements. M A single advertisement MUST be broken into separately sent components if there is more than one Certificate option, in order to avoid excessive fragmentation at the IP layer. Unlike the fragmentation at the IP layer, individual components of an advertisement may be stored and taken in use before all the components have arrived; this makes them slightly more reliable and less prone to Denial-of-Service attacks. The 'M' flag, when set, indicates that there are more components coming in this advertisement. Component This is a 15 bit unsigned integer field. The first message in a multi-component advertisement has the Component field set to 0, the second set to 1, and so on. Reserved This field is unused. It MUST be initialized to zero by the sender and MUST be ignored by the receiver. Arkko, et al. Expires August 25, 2003 [Page 18] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Valid Options: Certificate Zero or one certificates are provided in Certificate options, to establish a certificate chain to a trusted root. Trusted Root Zero or more Trusted Root options may be included to help receivers decide which advertisements are useful for them. If present, these options MUST appear in the first component of a multi-component advertisement. Future versions of this protocol may define new option types. Receivers MUST silently ignore any options they do not recognize and continue processing the message. 6.3 Trusted Root Option The format of the Trusted Root option is as described in the following: 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 | Name Type | Name Length | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Name ... +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Where the fields are as follows: Type TBD for Trusted Root. Length The length of the option (including the Type, Length, Name Type, Name Length, and Name fields) in units of 8 octets. Name Type The type of the name included in the Name field. This specification defines only one legal value for this field: Arkko, et al. Expires August 25, 2003 [Page 19] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 1 FQDN Name Length The length of the Name field, in bytes. Octets beyond this length but within the length specified by the Length field are padding and MUST be set to zero by senders and ignored by receivers. Name When the Name Type field is set to 1, the Name field contains the Fully Qualified Domain Name of the trusted root, for example "trustroot.operator.com". 6.4 Certificate Option The format of the certificate option is as described in the following: 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 | Cert Type | Pad Length | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Certificate ... +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Where the fields are as follows: Type TBD for Certificate. Length The length of the option (including the Type, Length, Cert Type, Pad Length, and Certificate fields) in units of 8 octets. Cert Type The type of the certificate included in the Name field. This specification defines only one legal value for this field: 1 X.509 Certificate Arkko, et al. Expires August 25, 2003 [Page 20] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Pad Length The amount of padding beyond the end of the Certificate field but within the length specified by the Length field. Padding MUST be set to zero by senders and ignored by receivers. Certificate When the Cert Type field is set to 1, the Certificate field contains an X.509 certificate [10]. 6.5 Processing Rules for Routers Routers SHOULD possess a keypair and certificate from at least one certificate authority. A router MUST silently discard any received Delegation Chain Solicitation messages that do not satisfy all of the following validity checks: o The IP Hop Limit field has a value of 255, i.e., the packet could not possibly have been forwarded by a router. o If the message includes an IP Authentication Header, the message authenticates correctly. o ICMP Checksum is valid. o ICMP Code is 0. o ICMP length (derived from the IP length) is 8 or more octets. o Identifier field is non-zero. o All included options have a length that is greater than zero. The contents of the Reserved field, and of any unrecognized options, MUST be ignored. Future, backward-compatible changes to the protocol may specify the contents of the Reserved field or add new options; backward-incompatible changes may use different Code values. The contents of any defined options that are not specified to be used with Router Solicitation messages MUST be ignored and the packet processed as normal. The only defined option that may appear is the Trusted Root option. A solicitation that passes the validity checks is called a "valid solicitation". Routers MAY send unsolicited Delegation Chain Advertisements for Arkko, et al. Expires August 25, 2003 [Page 21] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 their trusted root. When such advertisements are sent, their timing MUST follow the rules given for Router Advertisements in RFC 2461 [6]. The only defined option that may appear is the Certificate option. At least one such option MUST be present. Router SHOULD also include at least one Trusted Root option to indicate the trusted root on which the Certificate is based. In addition to sending periodic, unsolicited advertisements, a router sends advertisements in response to valid solicitations received on an advertising interface. A router MAY choose to unicast the response directly to the soliciting host's address (if the solicitation's source address is not the unspecified address), but the usual case is to multicast the response to the all-nodes group. In a solicited advertisement, the router SHOULD include suitable Certificate options so that a delegation chain to the solicited root can be established. The root is identified by the FQDN from the Trusted Root option being equal to an FQDN in the AltSubjectName field of the root's certificate. The router SHOULD include the Trusted Root option(s) in the advertisement for which the delegation chain was found. If the router is unable to find a chain to the requested root, it SHOULD send an advertisement without any certificates. In this case the router SHOULD include the Trusted Root options which were solicited. Rate limitation of Delegation Chain Advertisements is performed as specified for Router Advertisements in RFC 2461 [6]. 6.6 Processing Rules for Hosts Hosts SHOULD possess the certificate of at least one certificate authority, and MAY possess their own keypair and certificate from this authority. A host MUST silently discard any received Router Advertisement messages that do not satisfy all of the following validity checks: o IP Source Address is a link-local address. Routers must use their link-local address as the source for Router Advertisement and Redirect messages so that hosts can uniquely identify routers. o The IP Hop Limit field has a value of 255, i.e., the packet could not possibly have been forwarded by a router. o If the message includes an IP Authentication Header, the message authenticates correctly. Arkko, et al. Expires August 25, 2003 [Page 22] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 o ICMP Checksum is valid. o ICMP Code is 0. o ICMP length (derived from the IP length) is 16 or more octets. o All included options have a length that is greater than zero. The contents of the Reserved field, and of any unrecognized options, MUST be ignored. Future, backward-compatible changes to the protocol may specify the contents of the Reserved field or add new options; backward-incompatible changes may use different Code values. The contents of any defined options that are not specified to be used with Router Advertisement messages MUST be ignored and the packet processed as normal. The only defined option that may appear is the Certificate option. An advertisement that passes the validity checks is called a "valid advertisement". Hosts SHOULD store all certificates retrieved in Delegation Chain Advertisements for use in subsequent verification of Router (and optionally Neighbor) Advertisements. Note that it may be useful to cache this information and implied verification results for use over multiple attachments to the network. When an interface becomes enabled, a host may be unwilling to wait for the next unsolicited Delegation Chain Advertisement. To obtain such advertisements quickly, a host SHOULD transmit up to MAX_RTR_SOLICITATIONS Delegation Chain Solicitation messages each separated by at least RTR_SOLICITATION_INTERVAL seconds. Delegation Chain Solicitations may be sent after any of the following events: o The interface is initialized at system startup time. o The interface is reinitialized after a temporary interface failure or after being temporarily disabled by system management. o The system changes from being a router to being a host, by having its IP forwarding capability turned off by system management. o The host attaches to a link for the first time. o A movement has been indicated by lower layers or has been inferred from changed information in a Router Advertisement. o The host re-attaches to a link after being detached for some time. o A Router Advertisement has been received with a public key that is not stored in the hosts' cache of certificates, or there is no Arkko, et al. Expires August 25, 2003 [Page 23] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 authorization delegation chain to the host's trusted root. Delegation Chain Solicitations MUST NOT be sent if a valid certificate chain exists in the host's cache from the desired router (or host) to the host's trusted root. A host MUST send Delegation Chain Solicitations either to the All-Routers multicast address, if it hasn't selected a default router yet, or to the default router's IP address if it has already been selected. If two hosts communicate with the solicitations and advertisements, these MUST be unicast to the hosts's address. Delegation Chain Solicitations SHOULD be rate limited and timed similarly with Router Solicitations, as specified in RFC 2461 [6]. When processing a possible advertisement sent as a response to a solicitation, the host MAY prefer to process first those advertisements with the same Identifier field value as in the solicitation. This make Denial-of-Service attacks against the mechanism harder (see Section 12.2). Arkko, et al. Expires August 25, 2003 [Page 24] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 7. IPsec Extensions In order to use IPsec in securing Neighbor and Router Discovery some extensions have been specified in this document. These include a new transform suitable for the use of public keys and/or CGAs, a timestamp mechanism suitable for replay protection in a multicast environment, and some extensions to security association and security policy databases. 7.1 The AH_RSA_Sig Transform The AH_RSA_Sig transform specifies how AH can be used without a symmetric key. This transform introduces the use of a new reserved SPI number and a new format for the Authentication Data field in AH. AH_RSA_Sig MUST NOT be negotiated in IKE. For consistency it has an IPsec DOI [4] Transform ID TBD , however. 7.1.1 Reserved SPI Number The AH_RSA_Sig MUST be only be used with the reserved SPI number TBD . 7.1.2 Authentication Data Format The format of the Authentication Data field in AH depends on the chosen transform. For the AH_RSA_Sig transform, the format is as follows: Arkko, et al. Expires August 25, 2003 [Page 25] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | PK_Len | Nonce_Len | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | + Timestamp + | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | . . . Public key . . . | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | . . . Nonce (optional) . . . | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | | . . . Digital Signature (remaining bytes) . . . | | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ The meaning of the fields is described below: PK_Len This 16 bit unsigned integer field contains the length of the Public Key field in bytes. Nonce_Len This 16 bit unsigned integer field contains the length of the Nonce field in bytes. The length is set to zero if that field is not present. Timestamp This 64 bit unsigned integer field contains a timestamp used for replay protection (the Sequence Number field in AH is not used for AH_RSA_Sig). The use of this field is discussed in Section 7.1.4. Public key This variable length field contains the public key of the sender Arkko, et al. Expires August 25, 2003 [Page 26] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 in X.509 format [10]. Nonce This variable length field, if present, contains the nonce used in the construction of the CGA address. Digital Signatures This variable length field, if present, contains the signature made using the sender's private key, over the the whole packet as defined by the usual AH rules [3]. The signature is made using the RSA algorithm and MUST be encoded as private key encryption in PKCS #1 format [11]. 7.1.3 AH_RSA_Sig Security Associations Security associations that specify the use of AH_RSA_Sig transform MUST record the following additional configuration information: o A flag that indicates whether or not authorization delegation to a trusted root is used. o A flag that indicates whether or not CGA addresses are used. Incoming security associations MUST also record the following additional information: o The public key of the trusted root, if authorization delegation is in use. o The minimum acceptable key length for peer public keys (and any intermediaries between the trusted root and the peer). The default SHOULD be 768 bits. Implementations MAY also set an upper limit in order to limit the amount of computation they need to perform when verifying packets that use these security associations. o The minimum acceptable Sec value, if CGA verification is required. Outgoing security associations MUST also record the following additional information: o A public-private keypair. If authorization delegation is in use, there must exist a delegation chain from a trusted root to this keypair. Arkko, et al. Expires August 25, 2003 [Page 27] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 o Optionally any information required to construct CGA signatures, including the used Sec value and nonce, and the resulting CGA address. 7.1.4 Replay Protection For AH_RSA_Sig, the Sequence Number field in AH MUST be set to zero by the sender and ignored by receivers. If anti-replay has been enabled in the security association, senders MUST set the Timestamp field to the current time. The format is 64 bits, and the contents are the number of milliseconds since January 1, 1970 00:00 UTC. If anti-replay has been enabled, receivers MUST be configured with an allowed Delta value and maintain a cache of messages received within this time period from each specific source address. Receivers MUST then check the Timestamp field as follows: o A packet with a Timestamp field value beyond the current time plus or minus the allowed Delta value MUST be silently discarded. o A packet accepted according to the above rule MUST be checked for uniqueness within the cache of received messages from the given source address. A packet that has already been seen from the same source with the same Timestamp field value MUST be silently discard. o A packet that passes both of the above tests MUST be registered in the cache for the given source address. o If the cache becomes full, the receiver SHOULD temporarily reduce the Delta value for that source address so that all messages within that value can still be stored. 7.1.5 Processing Rules for Senders A node sending a packet using the AH_RSA_Sig transform MUST construct the packet as follows: o The Next Header, Payload Len, and Reserved fields are set as described in RFC 2402. o The Security Parameters Index is set to the value specified in Section 7.1.1. Arkko, et al. Expires August 25, 2003 [Page 28] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 o The Sequence Number field is set to 0. o The PK_Len field in Authentication Data is set to the length of the public key used for signing this packet. This public key is stored in the security association. The key itself is put to the Public key field. o If the security association has specified the use of the CGA method, the Nonce_len field is set to the length of the nonce used in the construction of the CGA address. In this case the nonce is copied to the Nonce field. Otherwise, the Nonce_Len field is set to zero and the Nonce field is omitted. o The Timestamp field is set as described in Section 7.1.4. o The packet, in the form defined for AH's coverage, is signed using the private key in the security association, and the resulting PCKS #1 signature is put to the Digital Signature field. o Additionally, if the use of CGA has been specified for the security association we require that the source address of the packet has been constructed as specified in Section 5. A sending node uses as inputs the sender's public key, nonce, the subnet prefix from the Target Address, and the Target Link Layer Address. The Target Address (including the subnet prefix) is put to the Source Address field in the IPv6 header, and the public key and the nonce are put to the Authentication Data field in the AH header. 7.1.6 Processing Rules for Receivers A packet received on a security association employing AH_RSA_Sig transform MUST be checked as follows: o Next Header and Payload Len fields are valid as specified in RFC 2402. o The SPI field is equal to the value defined in Section 7.1.1. o The sum of the PK_Len, Nonce_Len, and LLA_Len fields does not exceed the length of the Authentication Data field. o The Nonce_Len field is non-zero if the use of CGA has been specified in the security association. o The Timestamp field is verified as described in Section 7.1.4. Arkko, et al. Expires August 25, 2003 [Page 29] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 o If the use of CGA has been specified in the security association, we additionally require that A node receiving an Neighbor or Router Advertisement message with CGA protection first checks the CGA address in the Target Address field by generating the address using the algorithm described in Section 5.3. The inputs for the algorithm are the sender's public key and nonce, included in the AH packet as described in Section 7.1.2, the subnet prefix from the Target Address, the Target Link Layer Address, which MUST be included in a Target Link Layer Address option, and the security parameter from the rightmost two bits of the Target Address. If the interface identifier checks, the recipient proceeds with the cryptographically more time consuming check of the AH signature. Note that a receiver which does not support CGA or has not specified its use in its security associations can still verify packets using trusted roots, even if CGA had been used on a packet. The CGA property of the address is simply left untested. o The Public key and Digital Signature fields can be correctly decoded, and the the Digital Signature verifies as specified in the previous section. o If the use of a trusted root has been configured for the security association, a valid authorization delegation chain is known between the receiver's trusted root and the sender's public key. Note that the receiver may verify just the CGA property of a packet, even if the sender has used a trusted root as well. Packets that do not pass all the above tests MUST be silently discarded. 7.2 Other IPsec Extensions 7.2.1 Destination Agnostic Security Associations In order to allow the same security association to be used when the the node sends packets to different peers using the same addresses, a change must be provided to the RFC 2401 rules on how security associations are identified. This change is particularly important, for instance, when routers use the same keys and security association to send Router Advertisements for up to number of prefixes x 2^64 hosts on an interface. The change is mandatory for all nodes that support the AH_RSA_Sig transform. Security associations that use the SPI value specified in Section 7.1.1 MUST be identified solely by the SPI and protocol numbers, not by the destination IP address. Arkko, et al. Expires August 25, 2003 [Page 30] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 7.2.2 ICMP Type Specific Selectors In order to allow finer granularity of protection for various ICMPv6 messages, it is necessary to extend the security policy database and security association selectors with the capability to distinguish between different messages. All nodes that support the AH_RSA_Sig transform MUST be capable of using ICMP and ICMPv6 Type field as a selector. Arkko, et al. Expires August 25, 2003 [Page 31] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 8. Securing Neighbor Discovery with SEND This section describes how to use IPsec and the mechanisms from Section 5, Section 6, Section 7 in order to provide security for Neighbor Discovery. 8.1 Using IPsec to Secure Neighbor Advertisement Messages All Neighbor Solicitation messages SHOULD be sent without protection. All Neighbor Advertisement messages MUST be protected with IPsec, using the AH_RSA_Sig transform. The protection can be based on CGA addresses, node certificates and trusted roots, or both as specified in the security association. All nodes MUST have the necessary key pairs, and as applicable, certificates and CGA parameters associated with their relationship to trusted root or to an address. Hosts that use stateless address autoconfiguration MUST generate new CGA addresses as specified in Section 5 for each new autoconfiguration run. It is outside the scope of this specification to describe trusted roots and address autoconfiguration (stateful or stateless) with dynamically changing addresses works. It is also outside the scope of this specification to describe how stateful address autoconfiguration works with the CGA method. Hosts MAY use Authorization Delegation Discovery to learn the certificate chain of their default router or peer host. 8.2 Security Policy and SA Database Configuration This section gives a description for the security policy and security associations database entries, under which the outbound and inbound Neighbor Advertisement messages can be protected. The following table summarizes the inbound security policy data base along with the inbound security associations: Arkko, et al. Expires August 25, 2003 [Page 32] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Policy entries: *------------------------------------------------------------------* | Proto: Type | Source | Destination | Treatment | *------------------------------------------------------------------* | ICMPv6: NS | * | * | pass | *------------------------------------------------------------------* | ICMPv6: NA | * | own | SA = NA_In | *------------------------------------------------------------------* | ICMPv6: NA | * | all-nodes MC | SA = NA_In | *------------------------------------------------------------------* Security associations: +------------------------------------------------------------------+ | Name | Direction | SPI | Proto | Transform | +------------------------------------------------------------------+ | NA_In | Inbound | TBD (fixed) | AH | AH_RSA_Sig | | | | | | CGA = yes/no | | | | | | root = ... (opt)| +------------------------------------------------------------------+ The following table summarizes outbound security policy database: Policy entries: *------------------------------------------------------------------* | Proto: Type | Source | Destination | Treatment | *------------------------------------------------------------------* | ICMPv6: NS | * | * | pass | *------------------------------------------------------------------* | ICMPv6: NA | own | * | SA = NA_Out | *------------------------------------------------------------------* Security associations: +------------------------------------------------------------------+ | Name | Direction | SPI | Proto | Transform | +------------------------------------------------------------------+ | NA_Out | Outbound | TBD (fixed) | AH | AH_RSA_Sig | | | | | | key pair = ... | | | | | | CGA = yes/no | | | | | | CGA params = ...| | | | | | root = ... (opt)| +------------------------------------------------------------------+ Arkko, et al. Expires August 25, 2003 [Page 33] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 9. Securing Router Discovery with SEND This section describes how to use IPsec and the mechanisms from Section 5, Section 6, Section 7 in order to provide security for Router Discovery. 9.1 Using IPsec to Secure Router Advertisement Messages All Router Solicitation messages SHOULD be sent without protection. All Router Advertisement messages MUST be protected with IPsec, using the AH_RSA_Sig transform. The protection can be based on CGA addresses, node certificates and trusted roots, or both as specified in the security association. All routers MUST have the necessary key pairs, and as applicable, certificates and CGA parameters associated with their relationship to trusted root or to an address. All hosts MUST have the certificate of a trusted root. Hosts SHOULD use Authorization Delegation Discovery to learn the certificate chain of their default router or peer host. 9.2 Using IPsec to Secure Redirect Messages All Redirect messages MUST be protected with IPsec, using the AH_RSA_Sig transform. The protection can be based on CGA addresses, node certificates and trusted roots, or both as specified in the security association. If only CGA-based security associations are used, hosts MUST follow the rules defined below when receiving Redirect messages: 1. The Redirect message MUST be protected as discussed above. 2. The receiver MUST verify that the Redirect message comes from an IP address to which the host may have earlier sent the packet that the Redirect message now partially returns. That is, the source address of the Redirect message must be the default router for traffic sent to the destination of the returned packet. If this is not the case, the message MUST be silently discarded. This step prevents a bogus router from sending a Redirect message when the host is not using the bogus router as a default router. Arkko, et al. Expires August 25, 2003 [Page 34] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 9.3 Security Policy and SA Database Configuration This section gives a description for the security policy and security associations database entries, under which the outbound and inbound Router Advertisement and Redirect messages can be protected. The following table summarizes the inbound security policy data base along with the inbound security associations: Policy entries: *------------------------------------------------------------------* | Proto: Type | Source | Destination | Treatment | *------------------------------------------------------------------* | ICMPv6: RS | * | * | pass | *------------------------------------------------------------------* | ICMPv6: RA | * | own | SA = RA_In | *------------------------------------------------------------------* | ICMPv6: RA | * | all-nodes MC | SA = RA_In | *------------------------------------------------------------------* | ICMPv6: REDIRECT | * | own | SA = RE_In | *------------------------------------------------------------------* Security associations: +------------------------------------------------------------------+ | Name | Direction | SPI | Proto | Transform | +------------------------------------------------------------------+ | RA_In | Inbound | TBD (fixed) | AH | AH_RSA_Sig | | | | | | CGA = yes/no | | | | | | root = ... (opt)| +------------------------------------------------------------------+ | RE_In | Inbound | TBD (fixed) | AH | AH_RSA_Sig | | | | | | CGA = yes/no | | | | | | root = ... (opt)| +------------------------------------------------------------------+ The following table summarizes outbound security policy database. The Router Advertisement and Redirect entries are only present in routers. Arkko, et al. Expires August 25, 2003 [Page 35] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Policy entries: *------------------------------------------------------------------* | Proto: Type | Source | Destination | Treatment | *------------------------------------------------------------------* | ICMPv6: RS | * | * | pass | *------------------------------------------------------------------* | ICMPv6: RA | own | * | SA = RA_Out | *------------------------------------------------------------------* | ICMPv6: REDIRECT | own | * | SA = RE_Out | *------------------------------------------------------------------* Security associations: +------------------------------------------------------------------+ | Name | Direction | SPI | Proto | Transform | +------------------------------------------------------------------+ | RA_Out | Outbound | TBD (fixed) | AH | AH_RSA_Sig | | | | | | key pair = ... | | | | | | CGA = yes/no | | | | | | CGA params = ...| | | | | | root = ... (opt)| +------------------------------------------------------------------+ | RE_Out | Outbound | TBD (fixed) | AH | AH_RSA_Sig | | | | | | key pair = ... | | | | | | CGA = yes/no | | | | | | CGA params = ...| | | | | | root = ... (opt)| +------------------------------------------------------------------+ Arkko, et al. Expires August 25, 2003 [Page 36] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 10. Operational Considerations During the transition to secure links or as a policy consideration, network operators may want to run a particular link with a mixture of secure and insecure nodes. In such a case, the link is required to operate as two separate logical links, and packets between a secure and insecure node always go through the router. Routers configured for SEND advertise two sets of globally routable prefixes: one set for SEND nodes and one set for nodes that implement insecure Neighbor Discovery. The insecure nodes will ignore the advertisements sent using SEND, as the original Neighbor Discovery specifications require silently discarding packets if they contain an AH header that they can not verify. The following considerations apply to hosts: o Hosts configured for SEND MUST use SEND for all of their addresses, including link local addresses. o Hosts configured for SEND MUST validate all Router Advertisements with the protocol described in Section 8. Note that this includes discarding advertisements received without a valid IPsec AH header and CGA address, thus making insecure prefixes invisible to them. o Hosts configured for SEND MUST secure and validate all Neighbor Advertisements with the protocol described in Section 8. Note that this includes discarding advertisements received without a valid IPsec AH header and CGA address. The following considerations apply to routers: o Routers MUST send two sets of Router Advertisements. The advertisements containing the secure prefixes MUST be secured with the protocol described in Section 9. The advertisements containing the insecure prefixes MUST be sent without security. o Routers MUST assign different addresses for their secure and insecure communications, including their link-local addresses. Secure Router and Neighbor Advertisements MUST use a source address that satisfies the security properties outlined in Section 9. Unless this address is link-local, it MUST belong to one of the advertised secure prefixes. Similarly, source addresses for insecure advertisements MUST belong to one of the advertised insecure prefixes, unless the address is link-local. o Routers MUST refrain from sending Redirects to a SEND-secured node with the Destination Address field set to an address for an Arkko, et al. Expires August 25, 2003 [Page 37] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 insecure node. Similarly, routers MUST refrain from sending Redirects to a insecure node with the Destination Address field set to an address for a SEND-secured node The above rules require secure nodes to ignore all insecure Neighbor and Router Discovery messages, and all insecure nodes to ignore all SEND-secured messages. This implies that the secure and insecure nodes will not be able to discover each other, or even realize that the other prefixes are on-link. Thus, these hosts will request the router to route packets destined to the a host in the other group. The rules regarding Redirect messages above have been provided to ensure that the router performs its routing task and does not instruct the hosts to communicate directly. One effect of this is that secure hosts can not communicate with insecure hosts using link-local addresses, and vice versa. The security policy or security association database entries are needed for insecure nodes as far as Neighbor Discovery is concerned. SEND-secured nodes have the usual entries required by SEND. Arkko, et al. Expires August 25, 2003 [Page 38] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 11. Performance Considerations The computations related to AH_RSA_Sig transform are substantially more expensive than those with traditional symmetric transforms. While computational power is increasing, it appears still impractical to use asymmetric transforms for a significant amount packets. In the application for which AH_RSA_Sig has been designed, however, hosts typically have the need to perform only a few operations as they enter a link, and a few operations as they find a new on-link peer to communicate with. Routers are required to perform a larger amount of operations, particularly when the frequency of router advertisements is high due to mobility requirements. Still, the number of operations on a router is in the order of a few dozen operations per second, some of which can be precomputed as discussed below. A large number of router solicitations may cause higher demand for performing asymmetric operations, although RFC 2461 limits the rate at which responses to solicitations can be sent. Signatures related to the use of the AH_RSA_Sig transform MAY be precomputed for Multicast Neighbor and Router Advertisements. Typically, solicited advertisements are sent to the unicast address from which the solicitation was sent. Given that the IPv6 header is covered by the AH integrity protection, it is typically not possible to precompute solicited advertisements. Arkko, et al. Expires August 25, 2003 [Page 39] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 12. Security Considerations 12.1 Achieved Security Properties The CGA method assures that the received messages are coming from the owner of the address. However, this method does not eliminate all security vulnerabilities related to the ND functions. CGA prevents spoofed answers to DAD queries. An attacker may still be able to prevent valid responses or requests from reaching the intended recipient. As a result both participants are forced to believe that no address collision exists, when there in fact is. Within Address Resolution and NUD functions CGA can be used to prevent spoofed responses. However, it is still possible to prevent the Address Resolution and NUD from completing for a given address. For the NUD, this means that a node is claimed to be unreachable, when it really is not. Hosts can use CGA to show that the Redirect messages come from their current router. Still, we cannot say anything about the other router mentioned in the Redirect message. When trusted roots are used to certify routers, this is, however, not an issue. Within the Router Discovery functionality the CGA method ensures that we are communicating with the same router all the time, and prevents spoofing of the link-layer address of the router. But it does not help to verify that the router is connected to the Internet or that it is authorized to advertise a specific route prefix. A proper verification of these properties will not be possible without involving a trusted root. Protection of Router (or Neighbor) Discovery with trusted roots ensures that the given router (or neighbor) belongs to the set of trusted entities. It does not provide assurance that the given router is not spoofing another legitimate router (but see Section 14). 12.2 Attacks against SEND Itself The CGA addresses have a 60-bit hash. This length is in within the range of an feasible attack in the future. The following mechanisms have been built in this draft to counteract such attacks: The inclusion of the routing prefix prevents precomputation attacks. The Sec parameter helps the SEND algorithm to scale as Moore's law increases processing power. Additional amount of computational Arkko, et al. Expires August 25, 2003 [Page 40] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 effort is involved in for both attackers and owners of an address; verifiers of a message still need to spend the same amount of effort. Some Denial-of-Service attacks against ND and SEND itself remain. For instance, an attacker may try to produce a very high number of packets that a victim host or router has to verify using asymmetric methods. While safeguards are required to prevent an excessive use of resources, this can still render the SEND in-operational. Security associations based on the use of asymmetric cryptography can be vulnerable to Denial-of-Service attacks, particularly when the attacker can guess the SPIs and destination addresses used in the security associations. In SEND this is easy, as both the SPIs and the addresses (such as all nodes multicast address) are standardized. Due to the use of multicast, one packet sent by the attacker will be processed by multiple receivers. When CGA protection is used, SEND deals with these attacks using the verification process described Section 7.1.6. In this process a simple hash verification of the CGA property of the address is performed first before performing the more expensive signature verification. When trusted roots and certificates are used in SEND, the defenses are not quite as effective. Implementations SHOULD track how much resources are being devoted to the processing of packets received with the AH_RSA_Sig transform, and start selectively dropping packets if too much resources are spent. Implementations MAY also start first dropping packets that which are not protected with CGA. The Authorization Delegation Discovery process may also be vulnerable to Denial-of-Service attacks. An attack may target a router by request a large number of delegation chains to be discovered for different roots. Routers SHOULD defend against such attacks by caching discovered information (including negative responses) and by limiting the number of different discovery processes they engage in. Attackers may also target hosts by sending a large number of unnecessary certificate chains, forcing hosts to spend useless memory and verification resources for them. Hosts defend against such attacks by limiting the amount of resources devoted to the certificate chains and their verification. Hosts SHOULD also prioritize advertisements sent as a response to their requests over multicast advertisements. Arkko, et al. Expires August 25, 2003 [Page 41] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 13. IANA Considerations This document defines two new ICMP message types, used in Authorization Delegation Discovery. These messages must be assigned ICMPv6 type numbers from the informational message range: o The Delegation Chain Solicitation message, described in Section 6.1. o The Delegation Chain Advertisement message, described in Section 6.2. This document defines two new Neighbor Discovery [6] options, which must be assigned Option Type values within the option numbering space for Neighbor Discovery messages: o The Trusted Root option, described in Section 6.3. o The Certificate option, described in Section 6.4. This document defines a new reserved SPI number in the Reserved SPI range 1-255 [3]. This document defines a new IPSEC AH Transform Identifier for the IPsec DOI [4]. This identifier represents the AH_RSA_Sig transform from Section 7.1. This document defines a new name space for the Name Type field in the Trusted Root option. Future values of this field can be allocated using standards action [5]. Another new name space is allocated for the Cert Type field in the Certificate option. Future values of this field can be allocated using standards action [5]. Arkko, et al. Expires August 25, 2003 [Page 42] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 14. Conclusions and Remaining Work This draft documents ongoing work. The following areas are still being studied: o Protection of solicitations. There are no provisions yet for the protection of Address Resolution which takes place as a side-effect of Neighbor Solicitations. Similarly, the effects of Duplicate Address Detection probes on other nodes currently doing DAD have not been covered, as they too are carried by solicitations. o CGA detailed format and calculation formulas: The CGA formulas used in this document are from an early approach to the control of the security level in an environment with a constrained number of output bits. An advanced version of this approach will be published soon and appears interesting [21]. o Transition issues: Security policy and security association database entry examples are needed before the correctness of the approach outlined in Section 10 can be estimated. Also, the ability of hosts to simultaneously use SEND and insecure ND without a router. The ability of a non-SEND router to participate on a link with SEND-capable hosts and other routers. o The security considerations, achieved security properties, and the treatment of Denial-of-Service attacks on the SEND mechanisms themselves need further work. o The formats used to carry trusted root references, certificates, and public keys may change. o It is unclear at this time how, and if, router and neighbor protection based on trusted roots relates to addresses and prefixes. Is a router only certified to use a particular IP address, or to provide a particular prefix to the link? o It is unclear whether MLD [16] protection is needed or not. Arkko, et al. Expires August 25, 2003 [Page 43] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Normative References [1] Hinden, R. and S. Deering, "IP Version 6 Addressing Architecture", RFC 2373, July 1998. [2] Kent, S. and R. Atkinson, "Security Architecture for the Internet Protocol", RFC 2401, November 1998. [3] Kent, S. and R. Atkinson, "IP Authentication Header", RFC 2402, November 1998. [4] Piper, D., "The Internet IP Security Domain of Interpretation for ISAKMP", RFC 2407, November 1998. [5] Narten, T. and H. Alvestrand, "Guidelines for Writing an IANA Considerations Section in RFCs", BCP 26, RFC 2434, October 1998. [6] Narten, T., Nordmark, E. and W. Simpson, "Neighbor Discovery for IP Version 6 (IPv6)", RFC 2461, December 1998. [7] Thomson, S. and T. Narten, "IPv6 Stateless Address Autoconfiguration", RFC 2462, December 1998. [8] Conta, A. and S. Deering, "Internet Control Message Protocol (ICMPv6) for the Internet Protocol Version 6 (IPv6) Specification", RFC 2463, December 1998. [9] Narten, T. and R. Draves, "Privacy Extensions for Stateless Address Autoconfiguration in IPv6", RFC 3041, January 2001. [10] International Organization for Standardization, "The Directory - Authentication Framework", ISO Standard X.509, 2000. [11] RSA Laboratories, "RSA Encryption Standard, Version 1.5", PKCS 1, November 1993. [12] National Institute of Standards and Technology, "Secure Hash Standard", FIPS PUB 180-1, April 1995, . Arkko, et al. Expires August 25, 2003 [Page 44] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Informative References [13] Postel, J., "Internet Control Message Protocol", STD 5, RFC 792, September 1981. [14] Plummer, D., "Ethernet Address Resolution Protocol: Or converting network protocol addresses to 48.bit Ethernet address for transmission on Ethernet hardware", STD 37, RFC 826, November 1982. [15] Harkins, D. and D. Carrel, "The Internet Key Exchange (IKE)", RFC 2409, November 1998. [16] Deering, S., Fenner, W. and B. Haberman, "Multicast Listener Discovery (MLD) for IPv6", RFC 2710, October 1999. [17] Arkko, J., "Effects of ICMPv6 on IKE and IPsec Policies", draft-arkko-icmpv6-ike-effects-01 (work in progress), June 2002. [18] Arkko, J., "Manual SA Configuration for IPv6 Link Local Messages", draft-arkko-manual-icmpv6-sas-01 (work in progress), June 2002. [19] Droms, R., "Dynamic Host Configuration Protocol for IPv6 (DHCPv6)", draft-ietf-dhc-dhcpv6-28 (work in progress), November 2002. [20] Montenegro, G. and C. Castelluccia, "SUCV Identifiers and Addresses", draft-montenegro-sucv-03 (work in progress), July 2002. [21] Aura, T., "Cryptographically Generated Addresses (CGA)", draft-aura-cga-00.txt (work in progress), February 2003. [22] O'Shea, G. and M. Roe, "Child-proof Authentication for MIPv6", Computer Communications Review, April 2001. [23] Nikander, P., "Denial-of-Service, Address Ownership, and Early Authentication in the IPv6 World", Proceedings of the Cambridge Security Protocols Workshop, April 2001. [24] Arkko, J., Aura, T., Kempf, J., Mantyla, V., Nikander, P. and M. Roe, "Securing IPv6 Neighbor Discovery", Wireless Security Workshop, September 2002. [25] Montenegro, G. and C. Castelluccia, "Statistically Unique and Cryptographically Verifiable (SUCV) Identifiers and Addresses", Arkko, et al. Expires August 25, 2003 [Page 45] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 NDSS, February 2002. Authors' Addresses Jari Arkko Ericsson Jorvas 02420 Finland EMail: jari.arkko@ericsson.com James Kempf DoCoMo Communications Labs USA 181 Metro Drive San Jose, CA 94043 USA EMail: kempf@docomolabs-usa.com Bill Sommerfeld SUN Microsystems USA EMail: sommerfeld@east.sun.com Brian Zill Microsoft USA EMail: bzill@microsoft.com Arkko, et al. Expires August 25, 2003 [Page 46] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Appendix A. Contributors Steven Bellovin was the first to suggest the use of IPsec in this manner for the protection of Neighbor Discovery. Pekka Nikander and Vesa-Matti Mantyla were co-authors of an unpublished draft from which many of the details of this document have been inherited. The theoretical foundations of protecting Neighbor Discovery were laid out in a paper [24] where Tuomas Aura, Vesa-Matti Mantyla, Pekka Nikander, and Mike Roe were co-authors. Arkko, et al. Expires August 25, 2003 [Page 47] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Appendix B. Acknowledgements The authors would like to thank Erik Nordmark and Gabriel Montenegro for interesting discussions in this problem space. Arkko, et al. Expires August 25, 2003 [Page 48] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Appendix C. IPR Considerations The optional CGA part of SEND uses public keys and hashes to prove address ownership. Several IPR claims have been made about such methods. Arkko, et al. Expires August 25, 2003 [Page 49] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 Intellectual Property Statement The IETF takes no position regarding the validity or scope of any intellectual property or other rights that might be claimed to pertain to the implementation or use of the technology described in this document or the extent to which any license under such rights might or might not be available; neither does it represent that it has made any effort to identify any such rights. Information on the IETF's procedures with respect to rights in standards-track and standards-related documentation can be found in BCP-11. Copies of claims of rights made available for publication and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementors or users of this specification can be obtained from the IETF Secretariat. The IETF invites any interested party to bring to its attention any copyrights, patents or patent applications, or other proprietary rights which may cover technology that may be required to practice this standard. Please address the information to the IETF Executive Director. Full Copyright Statement Copyright (C) The Internet Society (2003). 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 English. The limited permissions granted above are perpetual and will not be revoked by the Internet Society or its successors or assignees. 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 Arkko, et al. Expires August 25, 2003 [Page 50] Internet-Draft SEcure Neighbor Discovery (SEND) February 2003 HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Acknowledgement Funding for the RFC Editor function is currently provided by the Internet Society. Arkko, et al. Expires August 25, 2003 [Page 51]