BEHAVE WG F. Gont Internet-Draft UTN/FRH Intended status: BCP P. Srisuresh Expires: April 29, 2010 Kazeon Systems, Inc. October 26, 2009 Security implications of Network Address Translators (NATs) draft-gont-behave-nat-security-02.txt Status of this Memo This Internet-Draft is submitted to IETF in full conformance with the provisions of BCP 78 and BCP 79. This document may not be modified, and derivative works of it may not be created, and it may not be published except as an Internet-Draft. 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 April 29, 2010. Copyright Notice Copyright (c) 2009 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 in effect on the date of publication of this document (http://trustee.ietf.org/license-info). Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Gont & Srisuresh Expires April 29, 2010 [Page 1] Internet-Draft NAT security implications October 2009 Abstract This document analyzes the security implications of Network Address Translators (NATs). It neither deprecates nor encourages the use of NATs, but rather aims to raise awareness about their security implications, and possible implementation approaches to improve their security. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4 2. Security Implications from IP fragmentation . . . . . . . . . 4 2.1. Fragment processing for inbound IP packets . . . . . . . . 4 2.2. Fragment processing on the outbound . . . . . . . . . . . 5 3. Port reservation . . . . . . . . . . . . . . . . . . . . . . . 6 4. Peer-to-peer Communication for the end hosts behind devices . 7 5. DHCP-Configured NATs . . . . . . . . . . . . . . . . . . . . . 7 5.1. DHCP-Configured NATs in a Multi-Level NAT deployments . . 7 5.2. DHCP-Configured NATs in a Remote Access VPN operation . . 8 6. Secure Transport for the end hosts behind NAT Devices . . . . 8 7. Security considerations arising from protocol header fields . 9 7.1. Internet Protocol version 4 (IPv4) header fields . . . . . 9 7.1.1. Version . . . . . . . . . . . . . . . . . . . . . . . 9 7.1.2. IHL . . . . . . . . . . . . . . . . . . . . . . . . . 9 7.1.3. Type of Service . . . . . . . . . . . . . . . . . . . 9 7.1.4. Total Length . . . . . . . . . . . . . . . . . . . . . 9 7.1.5. Identification . . . . . . . . . . . . . . . . . . . . 10 7.1.6. Flags . . . . . . . . . . . . . . . . . . . . . . . . 10 7.1.7. Fragment Offset . . . . . . . . . . . . . . . . . . . 10 7.1.8. Time to Live . . . . . . . . . . . . . . . . . . . . . 10 7.1.9. Protocol . . . . . . . . . . . . . . . . . . . . . . . 10 7.1.10. Header Checksum . . . . . . . . . . . . . . . . . . . 10 7.1.11. Source Address . . . . . . . . . . . . . . . . . . . . 10 7.1.12. Destination Address . . . . . . . . . . . . . . . . . 11 7.1.13. Options . . . . . . . . . . . . . . . . . . . . . . . 11 7.1.14. Padding . . . . . . . . . . . . . . . . . . . . . . . 11 7.2. Transmission Control Protocol (TCP) header fields . . . . 11 7.2.1. Source Port . . . . . . . . . . . . . . . . . . . . . 11 7.2.2. Destination Port . . . . . . . . . . . . . . . . . . . 11 7.2.3. Sequence Number . . . . . . . . . . . . . . . . . . . 11 7.2.4. Acknowledgment Number . . . . . . . . . . . . . . . . 12 7.2.5. Data Offset . . . . . . . . . . . . . . . . . . . . . 12 7.2.6. Reserved . . . . . . . . . . . . . . . . . . . . . . . 12 7.2.7. Flags . . . . . . . . . . . . . . . . . . . . . . . . 12 7.2.8. Window . . . . . . . . . . . . . . . . . . . . . . . . 12 7.2.9. Checksum . . . . . . . . . . . . . . . . . . . . . . . 12 7.2.10. Urgent Pointer . . . . . . . . . . . . . . . . . . . . 13 Gont & Srisuresh Expires April 29, 2010 [Page 2] Internet-Draft NAT security implications October 2009 7.2.11. Options . . . . . . . . . . . . . . . . . . . . . . . 13 7.2.12. Padding . . . . . . . . . . . . . . . . . . . . . . . 13 8. Security Considerations . . . . . . . . . . . . . . . . . . . 13 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 14 10. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 14 11. References . . . . . . . . . . . . . . . . . . . . . . . . . . 14 11.1. Normative References . . . . . . . . . . . . . . . . . . . 14 11.2. Informative References . . . . . . . . . . . . . . . . . . 15 Appendix A. Change log (to be removed before publication of the document as an RFC) . . . . . . . . . . . . . . . 16 A.1. Changes from draft-gont-behave-nat-security-01 . . . . . . 16 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 16 Gont & Srisuresh Expires April 29, 2010 [Page 3] Internet-Draft NAT security implications October 2009 1. Introduction This document analyzes the security implications of Network Address Translators (NATs). It neither deprecates nor encourages the use of NATs, but rather aims to raise awareness about their security implications, and possible implementation approaches to improve their security. Section 2 and Section 3 analyze the possible security implications of basic NAT functionality. Section 5 analyzes the possible security implications of DHCP-Configured NATs. Section 7 analyzes the possible security implications araising from the non-modification of protocol header fields by NATs. Note: the security implications of a NAT device due to it being a stateful device are not discussed in the current version of this document (but may be added in future revisions). For what is worth, many of these security implications are described in [RFC5382], [RFC4787] and [I-D.ietf-behave-nat-icmp]. 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]. 2. Security Implications from IP fragmentation 2.1. Fragment processing for inbound IP packets Routers in the network are able to forward fragmented IP packets just as they do any other non-fragmented IP packets because packet forwarding is based solely on looking up the destination IP address in the routing table and finding the largest prefix match to identify the next-hop to forward to. Routers do not need to retain any state pertaining to fragmented packets traversing them. A NAT device operates differently from a router in that the NAT device must find the matching NAT Session for an IP packet and perform NAT translation on the packet, prior to forwarding. NAT Session lookup requires the full 5-tuple of the IP datagram. Only the first fragment of the IP datagram contains the full-tuple. Subsequent fragmented packets contain the fragment Id, but do not contain transport protocol specific details such as source and destination port numbers. The NAT device must be able to associate the same session tuple for all fragments by virtue of the fragment ID and use that information to locate the NAT Session the packets belong to. Note however, the IP fragments cannot be assumed to arrive in order. Some operating systems transmit the fragments of an IP Gont & Srisuresh Expires April 29, 2010 [Page 4] Internet-Draft NAT security implications October 2009 datagram out of their logical order as a matter of course. In addition, network conditions can also cause dynamic packet reordering in transit. A NAT device not capable of processing all fragments of an inbound IP datagram can cause the fragmented packets to be dropped causing some applications to not function correctly. NATs, capable of processing out-of-order packets store the out-of- order packets prior to forwarding. This can open up the NAT device for external attacks. As pointed out in [RFC4787], fragmentation has been a tool used in many attacks, some involving passing fragmented packets through NATs, and others involving DoS attacks based on the state needed to reassemble the fragments. NAT implementers should be aware of [RFC3128] and [RFC1858]. NATs may protect themselves against such attacks by limiting the length of time they retain an incomplete IP packet before discarding it, or by limiting the amount of internal buffer space incomplete IP packets may consume before the oldest fragments are discarded. The appropriate values of these limits vary across NATs, and may be determined by the network administrator. [CPNI-IP] contains a detailed discussion of the security implications arising from the reassembly of IP fragments and of a number of approaches to mitigate them. REQ-1: A NAT device capable of forwarding out-of-order IP fragments MUST take measures to protect itself against well-known IP fragment based attacks. 2.2. Fragment processing on the outbound Say, two private hosts originated TCP/UDP packets (fragmented or not) to the same destination host and both packets transit the same NAPT device and use the same fragmentation identifier. Say, the NAPT device assembled the IP packets (in the case they were fragmented) and translated the same using the appropriate NAT Sessions. When NAPT translates IP datagrams, it would assign all outbound IP datagrams the same Public IP, but different TCP/UDP numbers. While forwarding, an IP datagram may be fragmented on the way out. Only the first fragment contains the TCP/UDP header that would be necessary to associate the packet to a specific session. Subsequent fragments do not contain TCP/UDP port information, but simply carry the same fragmentation identifier specified in the first fragment. When the target host receives the two unrelated datagrams, carrying same fragmentation id, and from the same assigned host address, the Gont & Srisuresh Expires April 29, 2010 [Page 5] Internet-Draft NAT security implications October 2009 target host is unable to determine which of the two sessions the datagrams belong to and might corrupt both sessions. This can be a security breach any of the sessions associated. In order to avoid problems of this kind, the NAPT device MUST further translate fragment ID in the outgoing packets such that the tuples of (SrcIP, DestIP, fragment Id, Protocol) are unique and distinguishable across all outgoing packets from the NAT device. When fragmenting packets on the outbound, a NAT device implementing NAPT function MUST ensure that the tuples of (SrcIP, DestIP, fragment Id, Protocol) are unique across all outgoing packets. REQ-2: when forwarding IP fragments on the outbound, a NAT device implementing NAPT function MUST ensure that the tuple of (SrcIP, DestIP, Protocol, fragment ID) is unique across NAT Sessions. Doing this will eliminate the possibility of cross-session contamination due to fragment Identifier overlap when more than one NAT session share the same DestIP and Protocol. 3. Port reservation A NAT device implementing NAPT function shares the source ports for its public IP address with nodes in the private realm. The NAPT device may also have end host applications of its own. Consider the following scenario when a NAPT device uses the same TCP/UDP port for local use as well as for mapping to a private host. Say, an application on the NAPT device runs on port 5060 (SIP Server), but not enabled. Say, a host in the private domain of the Nat device also uses 5060 and obtains port 5060 from the NAT device. While this Port Binding is active, say, the application on NAPT device is activated. The application on the NAT device is unlikely to be aware of the NAT function enforced by the NAT device. Once the same port is assigned for NAT use as well as for use by local application on the NAT device, data packets directed to the NAT device could end up with the end host within the private domain or the application on the NAT device. This behavior can cause unpredictable behavior and may even result in data snooping. Manual intervention becomes necessary to ensure that only one instance of an application is actively using a port at a given time. It is not desirable to either allow possible simulataneous use (or) require manual intervention to serialize the use. REQ-3: When a NAT device supports local applications on the device, it is RECOMMENDED that the NAT device reserves specific ports for Gont & Srisuresh Expires April 29, 2010 [Page 6] Internet-Draft NAT security implications October 2009 local use, different from NAT use, so there is no overlap of ports between local use and NAT use. Doing this will ensure there is no possibility of cross session contamination between NAT sessions and local sessions. 4. Peer-to-peer Communication for the end hosts behind devices [RFC5128] refers to applications using TCP/UDP hole punching technique to establish peer-to-peer communication. Section 6.1 of [RFC5128] describes a scenario in which it can be problematic for applications that do not use appropriate authentication mechanisms while setting up peer connections. An application could end up connecting to the wrong host or have its connections hijacked maliciously by other hosts. REQ-4: Applications attempting to establish peer-to-peer communication across NAT devices using TCP/UDP hole punching technique SHOULD employ relevant authentication mechanism to connect to their peers. 5. DHCP-Configured NATs 5.1. DHCP-Configured NATs in a Multi-Level NAT deployments Many NATs, particularly consumer-level devices designed to be deployed by nontechnical users, also act as DHCP [RFC2131] clients. In its default configuration, a consumer NAT typically obtains its public IP address, default router, and other IP configuration information Via DHCP from an ISP or other "upstream" network. On its internal network side, the NAT then automatically sets up its own private "downstream" subnet in one of the private IP address regions assigned to this purpose in [RFC1918]. The NAT typically acts as a DHCP server for its private downstream network, managing its pool of private IP addresses automatically and handing them out to the hosts (and perhaps other NATs) on the private network on demand. This auto-configuration of private networks can be problematic, if the NAT's upstream network is also in RFC 1918 private address space. In the Multi Level NAT deployments, end-hosts could have their security compromised due to mistaken serveri identities as described in section 3 of [I-D.ford-behave-top]. REQ-5: DHCP-Configured NATs SHOULD use different address spaces for the private and the external realms. Gont & Srisuresh Expires April 29, 2010 [Page 7] Internet-Draft NAT security implications October 2009 5.2. DHCP-Configured NATs in a Remote Access VPN operation In deployments where a corporate network deploys the same private address space as used on sundry remote locations, end-hosts could have their security compromised due to mistaken server identities, as described in section 4 of [I-D.ford-behave-top]. 6. Secure Transport for the end hosts behind NAT Devices NAT devices are ubiquitous in the Internet. NAT devices can be found in homes, hotels, Airports, conferences, coffee shops and plethora of internet cafes. Most users needing to carry out financial transactions and other personal, sensitive applications use SSL/TLS protocol [RFC2246] to do this. NAT devices enroute MUST support the traversal of SSL/TLS protocol. TCP port 443 is the default port used for SSL/TLS protocol. Secure VPNs is another important use of secure protocols to access corporate networks. Telecommuters and users in remote locations use secure VPN to access their corporate networks. The secure VPNs may use a combination of NAT-T [RFC3947] and IPSec-over-UDP [RFC3948] to secure the VPN traffic. Alternately, some VPN vendors use SSL/TLS protocol [RFC2246] to secure the VPN traffic. NAT devices enroute MUST support the traversal of NAT compliant security protocols such as SSL/TLS, NAT-T and IPsec-over-UDP. Enforcement of NAT-T for IKE negotiation can be problematic, as described in Section 2.3 of [RFC3715], if the NAT device enroute has an Application Level Gateway (ALG) that attempts to treat IKE packets differently from other UDP packets. A NAT device MUST NOT include an ALG that treats IKE or SSL/TLS packets differently than any other TCP/UDP packet. REQ-6: A NAT device MUST permit the traversal of NAT compliant security protocols. Specifically, a NAT device MUST do the following. a. A NAT device MUST NOT block traffic directed to or coming from UDP port numbers 500 and 4500. b. A NAT device MUST NOT block traffic directed to or coming from TCP/UDP port number 443. c. A NAT device MUST NOT include an ALG that treats IKE packets or SSL/TLS packets differently than any other TCP/UDP packet. Gont & Srisuresh Expires April 29, 2010 [Page 8] Internet-Draft NAT security implications October 2009 7. Security considerations arising from protocol header fields From the external real, all packets originated by any host in the internal realm (or by the NAT itself) are seen as originating from the NAT box. As a result, the same requirements that are applied to an internet host must be applied to the aggregate traffic at the NAT box. For example, in the same way that an internet host is required not to reuse a tuple (SrcIP, DstIP, Protocol, IP ID) at any given time, a NAT should enforce that a tuple (SrcIP, DstIP, Protocol, IP ID) of the aggregate traffic is not reused at any given time. In order to enforce some of these requirements, a NAT will usually need to rewrite some of the TCP and/or IP header fields of the incoming and outgoing packets. The following subsections analyze the interoperability and security implications arising from the non-modification of protocol header fields by Network Address Translators (NATs) 7.1. Internet Protocol version 4 (IPv4) header fields 7.1.1. Version This field does not require any special handling by NATs. 7.1.2. IHL This field does not require any special handling by NATs. 7.1.3. Type of Service End-systems have traditionally selected different TOS values depending on the nature of the application making use of the IP services. As the TOS value selected for each packet usually depends the specific IP implementation, this could be exploited to roughly count the number of systems behind a NAT. In order to avoid the TOS field from revealing this information, NATs could rewrite the TOS field in outgoing packets according to the Protocol value in the IP header, and the Destination Port value in the header of the transport protocol running on top of IP. 7.1.4. Total Length This field does not require any special handling by NATs. Gont & Srisuresh Expires April 29, 2010 [Page 9] Internet-Draft NAT security implications October 2009 7.1.5. Identification NATs must ensure that a tuple (SrcIP, DstIP, Protocol, IP ID) is not reused while there are still packets in the network with that tuple. In order to enforce this requirement, NATs MUST rewrite the IP Identification field of the outgoing IP packets. A trivial approach to enforce this requirement would be to rewrite the IP Identification from a global counter that is increment by one each time a packet is transmitted. However, while this would fulfil the interoperability requirements, this would lead to predictable Identification values, which have been found to have a number of security implications [CPNI-IP]. In order to mitigate these security implications, NATs should rewrite the IP Identification field such that it is not trivial for an attacker to detect different "sequences" of the Identification field. [CPNI-IP] discusses a number of approaches for selecting the Identification value at end-systems, which could also be applied for the selection of the Identification value at NATs. REQ-6: NATs MUST ensure that a tuple (SrcIP, DstIP, Protocol, IP ID) is not reused while there are still packets in the network with that tuple. Additionally, they SHOULD generate the IP Identification values such that they are not trivially predictable. 7.1.6. Flags 7.1.7. Fragment Offset This field does not require any special handling by NATs. 7.1.8. Time to Live To be added. 7.1.9. Protocol This field does not require any special handling by NATs. 7.1.10. Header Checksum This field does not require any special handling by NATs. 7.1.11. Source Address Here we make no special considerations about this field. Gont & Srisuresh Expires April 29, 2010 [Page 10] Internet-Draft NAT security implications October 2009 7.1.12. Destination Address Here we make no special considerations about this field 7.1.13. Options Clearly, IP options could potentially be used for counting the number of systems behind a NAT. However, as it is unusual for end-systems to include IP options in the IP packets they send, in most cases this IP options will not require any special handling by NATs. 7.1.14. Padding This field does not require any special handling by NATs. 7.2. Transmission Control Protocol (TCP) header fields 7.2.1. Source Port As part of its basic functionality, a NAT-PT will usually rewrite (translate) the TCP Source Port of packets sent to the external realm. As a result, the ephemeral port selection algorithm of a NAT will "override" that of the end-systems behind the NAT. In some cases, this may have the undesireable consequence that a system implementing some algorithm for ephemeral port obfuscation may end up establishing TCP connections with systems in the external realm using a predictable (as seen from the external realm) ephemeral port sequence. NATs should implement an ephemeral port selection algorithm such that the source port of outgoing packets is obfuscated, thus mitigating blind (off-path) spoofing attacks. It should be noted that use of an improper ephemeral port selection algorithm may lead to collisions of connection-ids, with the potential of failure in the establishment of new TCP connections. [I-D.ietf-tsvwg-port-randomization] 7.2.2. Destination Port Here we make no special considerations for this field. 7.2.3. Sequence Number The choice of the Initial Sequence Number of a connection by an end- system is not arbitrary, but aims to minimize the chances of a stale segment from being accepted by a new incarnation of a previous Gont & Srisuresh Expires April 29, 2010 [Page 11] Internet-Draft NAT security implications October 2009 connection. [RFC0793] suggests the use of a global 32-bit ISN generator, whose lower bit is incremented roughly every 4 microseconds. Based on the premise that the ISNs of consecutive TCP connections are monotonically-increasing, BSD-derived implementations use the ISN of an incomming connection request to perform heuristics aiming at allowing a new incarnation of a previous connection to be created, even if the pervious incarnation is still in the TIME-WAIT state. However, an ISN such as that described in [RFC0793] makes it trivial for an attacker to predict the ISN used for future connections, thus making it easier for the attacker to perform "blind" attacks against those connections. NATs may rewrite the Sequence Number of outgoing segments such that consecutive connections to a specific service at a specific system use ISNs that are monotonically-increasing. Additionally, the ISN generator should be such that it should be difficult for an off-path attacker to predict the ISNs of future connections. [RFC1948] describes an algorithm for the generation of ISN that complies with these two "requirements". 7.2.4. Acknowledgment Number If the NAT rewrites the Sequence Number of TCP segments forwarded from the internal realm to the external realm, it must also rewrite the Acknowledgement Number of TCP segments forwarded from the external realm to the internal realm. 7.2.5. Data Offset Here we make no special considerations for this field. 7.2.6. Reserved Here we make no special considerations for this field. 7.2.7. Flags Here we make no special considerations for this field. 7.2.8. Window Here we make no special considerations for this field. 7.2.9. Checksum Here we make no special considerations for this field. Gont & Srisuresh Expires April 29, 2010 [Page 12] Internet-Draft NAT security implications October 2009 7.2.10. Urgent Pointer Here we make no special considerations for this field. 7.2.11. Options 7.2.11.1. TCP timestamps The Timestamps option, specified in RFC 1323 [RFC1323], allows a TCP to include a timestamp value in its segments, that can be used used to perform two functions: Round-Trip Time Measurement (RTTM), and Protect Against Wrapped Sequences (PAWS). For the purpose of PAWS, the timestamps sent on a connection are required to be monotonically increasing. While there is no requirement that timestamps are monotonically increasing across TCP connections, the generation of timestamps such that they are monotonically increasing across connections between the same two endpoints allows the use of timestamps for improving the handling of SYN segments that are received while the corresponding four-tuple is in the TIME-WAIT state. That is, the timestamp option could be used to perform heuristics to determine whether to allow the creation of a new incarnation of a connection that is in the TIME-WAIT state. NATs could rewrite the TCP Timestamps option such that TCP consecutive connections connections with a specific service at a specific system use monotonically increasing timestamps (i.e., the TCP timestamps are monotonically-increasing across those connections). [I-D.gont-tcpm-tcp-timestamps] describes an algorithm that complies with this requirement. 7.2.12. Padding Here we make no special considerations for this field. 8. Security Considerations This document analyzes the security implications of Network Address Translators (NATs). It neither deprecates nor encourages the use of NATs, but rather aims to raise awareness about their security implications, and possible implementation approaches to improve their security. Note: the security implications of a NAT device due to it being a stateful device are not discussed in the current version of this document (but may be added in future revisions). For what is worth, many of these security implications are described in [RFC5382], Gont & Srisuresh Expires April 29, 2010 [Page 13] Internet-Draft NAT security implications October 2009 [RFC4787] and [I-D.ietf-behave-nat-icmp]. 9. IANA Considerations This document has no actions for IANA. 10. Acknowledgements The authors of this document would like to thank (in alphabetical order) Brian Carpenter, Remi Denis-Courmont, Reinaldo Penno, Dave Thaler, and Dan Wing for providing valuable feedback on earlier versions of this document. 11. References 11.1. Normative References [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, September 1981. [RFC0793] Postel, J., "Transmission Control Protocol", STD 7, RFC 793, September 1981. [RFC1122] Braden, R., "Requirements for Internet Hosts - Communication Layers", STD 3, RFC 1122, October 1989. [RFC1323] Jacobson, V., Braden, B., and D. Borman, "TCP Extensions for High Performance", RFC 1323, May 1992. [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [RFC2246] Dierks, T. and C. Allen, "The TLS Protocol Version 1.0", RFC 2246, January 1999. [RFC3715] Aboba, B. and W. Dixon, "IPsec-Network Address Translation (NAT) Compatibility Requirements", RFC 3715, March 2004. [RFC3947] Kivinen, T., Swander, B., Huttunen, A., and V. Volpe, "Negotiation of NAT-Traversal in the IKE", RFC 3947, January 2005. [RFC3948] Huttunen, A., Swander, B., Volpe, V., DiBurro, L., and M. Stenberg, "UDP Encapsulation of IPsec ESP Packets", RFC 3948, January 2005. Gont & Srisuresh Expires April 29, 2010 [Page 14] Internet-Draft NAT security implications October 2009 11.2. Informative References [Bellovin2002] Bellovin, S., "A Technique for Counting NATted Hosts", IMW'02 Nov. 6-8, 2002, Marseille, France, 2002. [CPNI-IP] CPNI, "Security Assessment of the Internet Protocol", 2008 . [CPNI-TCP] CPNI, "Security Assessment of the Transmission Control Protocol (TCP)", (to be published) . [I-D.ford-behave-top] Srisuresh, P. and B. Ford, "Unintended Consequence of two NAT deployments with Overlapping Address Space", draft-ford-behave-top-07 (work in progress), August 2009. [I-D.gont-tcpm-tcp-timestamps] Gont, F., "On the generation of TCP timestamps", draft-gont-tcpm-tcp-timestamps-02 (work in progress), September 2009. [I-D.ietf-behave-nat-icmp] Srisuresh, P., Ford, B., Sivakumar, S., and S. Guha, "NAT Behavioral Requirements for ICMP protocol", draft-ietf-behave-nat-icmp-12 (work in progress), January 2009. [I-D.ietf-tsvwg-port-randomization] Larsen, M. and F. Gont, "Port Randomization", draft-ietf-tsvwg-port-randomization-04 (work in progress), July 2009. [RFC1858] Ziemba, G., Reed, D., and P. Traina, "Security Considerations for IP Fragment Filtering", RFC 1858, October 1995. [RFC1918] Rekhter, Y., Moskowitz, R., Karrenberg, D., Groot, G., and E. Lear, "Address Allocation for Private Internets", BCP 5, RFC 1918, February 1996. [RFC1948] Bellovin, S., "Defending Against Sequence Number Attacks", RFC 1948, May 1996. [RFC2131] Droms, R., "Dynamic Host Configuration Protocol", RFC 2131, March 1997. Gont & Srisuresh Expires April 29, 2010 [Page 15] Internet-Draft NAT security implications October 2009 [RFC2663] Srisuresh, P. and M. Holdrege, "IP Network Address Translator (NAT) Terminology and Considerations", RFC 2663, August 1999. [RFC3022] Srisuresh, P. and K. Egevang, "Traditional IP Network Address Translator (Traditional NAT)", RFC 3022, January 2001. [RFC3128] Miller, I., "Protection Against a Variant of the Tiny Fragment Attack (RFC 1858)", RFC 3128, June 2001. [RFC4787] Audet, F. and C. Jennings, "Network Address Translation (NAT) Behavioral Requirements for Unicast UDP", BCP 127, RFC 4787, January 2007. [RFC5128] Srisuresh, P., Ford, B., and D. Kegel, "State of Peer-to- Peer (P2P) Communication across Network Address Translators (NATs)", RFC 5128, March 2008. [RFC5382] Guha, S., Biswas, K., Ford, B., Sivakumar, S., and P. Srisuresh, "NAT Behavioral Requirements for TCP", BCP 142, RFC 5382, October 2008. Appendix A. Change log (to be removed before publication of the document as an RFC) A.1. Changes from draft-gont-behave-nat-security-01 o Addded Section 6. Authors' Addresses Fernando Gont Universidad Tecnologica Nacional / Facultad Regional Haedo Evaristo Carriego 2644 Haedo, Provincia de Buenos Aires 1706 Argentina Phone: +54 11 4650 8472 Email: fernando@gont.com.ar URI: http://www.gont.com.ar Gont & Srisuresh Expires April 29, 2010 [Page 16] Internet-Draft NAT security implications October 2009 Pyda Srisuresh Kazeon Systems, Inc. 1161 San Antonio Rd. Mountain View, CA 94043 U.S.A. Phone: +1 408 836 4773 Email: srisuresh@yahoo.com Gont & Srisuresh Expires April 29, 2010 [Page 17]