PCP Authentication Requirements
draft-reddy-pcp-auth-req-02

Abstract

In an attempt to reach consensus on a PCP authentication mechanism, this document describes requirements for PCP authentication. It is hoped this can serve as the basis for a comparison of PCP authentication mechanisms.

Status of This Memo

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at http://datatracker.ietf.org/drafts/current/.

Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."

This Internet-Draft will expire on October 24, 2013.

Copyright Notice

Copyright (c) 2013 IETF Trust and the persons identified as the document authors. All rights reserved.

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.

Table of Contents

• 1. Introduction
• 2. Terminology
• 3. Requirements
• 4. Third Party Authorization
• 5. Other recommendations
• 6. IANA Considerations
• 7. Security Considerations
• 8. References
• 8.1. Normative References
• 8.2. Informative References
• Appendix A. Change History
• A.1. Change from -01 to -02
• Authors' Addresses

1. Introduction

This document derives requirements for PCP Authentication from PCP deployment scenarios and scope described in PCP-base [I-D.ietf-pcp-base] and other PCP drafts. The document focuses on requirements and does not make a suggestion on the authentication mechanism to be used to satisfy requirements.

2. Terminology

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 [RFC2119].

3. Requirements

      +------------+                       |
      | PCP Client |-----+                 |
      +--(Host 1)--+     |   +-----------+ |     +----------+
                         +---|           | |     |          |
                             | PCP Proxy |-------|PCP Server|
                         +---|           | |     |          |
      +------------+     |   +-----------+ |     +----------+
      | PCP Client |-----+                 |
      +--(Host 2)--+               possible boundary
                              <- Home side | ISP side ->

REQ-1:

PCP client and server MUST provide client authentication. The client could be a host running a PCP client or middle box (e.g., NAT) running a PCP Proxy.

• The identity details of the client could be used by the PCP server to grant access to certain PCP opcodes or PCP options. For example GUESTS would not be permitted to use MAP opcode, ADMINISTRATOR is only permitted to use THIRD_PARTY option.
• The identity details of the client could be used for auditing.

PCP Authentication MUST also generate message authentication key for integrity protection of PCP request and response.

REQ-2:

PCP Servers MUST be able to indicate that a request will not be processed without authentication.

REQ-3:

If the original PCP request/response was authenticated,

1. A client MUST be able to verify the integrity and origin of a subsequent response from the server.
2. A server MUST be able to send subsequent authenticated unsolicited responses.
3. If a server wants to send an unsolicited message, but the previous security association has expired
   1. The server can continue to use the same SA to protect messages pertaining to that mapping, even if the SA is technically expired.
      • Such server notifications will not change state in the PCP client.
      • The notification could be a trigger for the client to re-authenticate. For example, if the server indicates that external IP address/port has changed, the PCP client can then re-authenticate with the server to confirm if the external IP address/port for the mapping has indeed changed.
   2. The server can optionally trigger re-authentication with the client.
4. If a PCP response does not include integrity related to a current security association, then those messages MUST NOT be trusted without soliciting an integrity protected version.

REQ-4:

It is important that PCP not leak privacy information between the PCP client and PCP server,

1. The authentication mechanism MUST be able to keep credentials hidden from eavesdroppers on path between client and server.
2. Confidentiality of the PCP messages is OPTIONAL for PCP request and response of opcodes MAP, PEER, ANNOUNCE and options THIRD_PARTY, PREFER_FAILURE and FILTER explained in PCP-base [I-D.ietf-pcp-base]. Other PCP drafts MUST evaluate if confidentiality is OPTIONAL or not for new PCP opcodes and options introduced.
3. PCP authentication SHOULD be immune to passive dictionary attacks.
4. PCP Authentication MUST ensure that an attacker snooping PCP messages cannot guess the SA.

REQ-5:

To ease troubleshooting and ensure fate sharing, PCP authentication and PCP messages MUST be multiplexed over the same port.

REQ-6:

PCP authentication MUST accommodate authentication between administrative domains. For example, a PCP client may wish to communicate directly to an ISP’s PCP server, even though the in-home CPE router does not support PCP. In this scenario the PCP client needs to directly authenticate with the ISP’s PCP server.

REQ-7:

PCP client and server MUST be able mutually authenticate, especially when the PCP server is located in a different administrative domain from the PCP client. Credentials to gain access to the network could be different from the credentials used to authenticate with the PCP server.

REQ-8:

For the scenarios described in REQ-6, PCP authentication mechanism MUST be functional across address and port translation, including NAPT64 and NAPT44.

REQ-9:

A PCP proxy, that modifies PCP request/response before forwarding messages,

1. MUST validate message integrity of PCP messages from the PCP server and client respectively.
2. MUST ensure message integrity after updating the PCP message for cases described in sections 6 and 7 of [I-D.ietf-pcp-proxy].

REQ-10:

It is RECOMMENDED that PCP authentication support a mechanism where only one PCP client on the host authenticates with the PCP server and other PCP clients be able to reuse the previously negotiated key for integrity protection. For example, multiple applications on the host like BitTorrent [BitTorrent], WebRTC[I-D.ietf-rtcweb-overview]/SIP [RFC3261] using PCP. Multiple authentication exchanges increase load on the PCP server and chatter on the network. For example, if 'N' messages are to be exchanged for PCP authentication and 'M' independent applications implement their own PCP client, a total of N*M messages have to be exchanged and 'M' number of SAs maintained for each host.

REQ-11:

All else equal, it is RECOMMENDED to choose a widely deployed authentication technique with known security properties rather than inventing a new authentication mechanism.

REQ-12:

Changes in PCP to accommodate authentication SHOULD be minimal so that updates and additions to the authentication mechanism have no bearing on modifying PCP.

4. Third Party Authorization

In addition to two party authentication that has been discussed in this draft, a mechanism for third party authorization must also be supported. This is required in cases where a third party authorizes the use of a resource on a PCP server for a desired PCP client. For example, a PCP request to a PCP capable firewall authorized by a SIP proxy rather than by virtue of the end user making the PCP request. The PCP server is to permit a PCP MAP request if a user is making a SIP call with the Enterprise SIP server, otherwise do not allow MAP request from that particular user. In this scenario the first party is the user, second party is the PCP server (which is also the firewall) and the third party is the SIP Server, where the user is authorized to use MAP request only when making a call using the trusted SIP Server.

5. Other recommendations

• It is recommended that there be support for a means to provide integrity protection without user authentication. For example, upon receiving a challenge with a certain REALM, if the PCP client does not have credentials for that REALM, the client will attempt to use a default username and password. Default credentials are expected to be configured on infrastructure where PCP authentication is not necessary, but such guest users are given some (minimal) authorization to use PCP. This addresses the problem when the client is visiting foreign networks like a hotel, hot spot etc where it may gain access to the network but does not know the credentials to authenticate to the ISP’s PCP server when the in-home CPE router does not support PCP and the PCP client needs to directly authenticate with the ISP’s PCP server.

6. IANA Considerations

This document does not require any action from IANA.

7. Security Considerations

This document does not define an architecture nor a protocol; as such it does not raise any security concerns.

8. References

8.1. Normative References

8.2. Informative References

Appendix A. Change History

A.1. Change from -01 to -02

• Requirements reorganized based on commonality
• New requirement 3(c(2)) added.

Authors' Addresses