<?xml version='1.0' encoding='utf-8'?>
<!DOCTYPE rfc [
  <!ENTITY nbsp    "&#160;">
  <!ENTITY zwsp   "&#8203;">
  <!ENTITY nbhy   "&#8209;">
  <!ENTITY wj     "&#8288;">
]>
<?xml-stylesheet type="text/xsl" href="rfc2629.xslt" ?>
<!-- generated by https://github.com/cabo/kramdown-rfc version 1.7.29 (Ruby 3.2.3) -->
<rfc xmlns:xi="http://www.w3.org/2001/XInclude" ipr="trust200902" docName="draft-geng-sidrops-aspa-analysis-03" category="info" submissionType="IETF" version="3">
  <!-- xml2rfc v2v3 conversion 3.30.0 -->
  <front>
    <title abbrev="ASPA analysis">An Analysis of ASPA-based AS_PATH Verification</title>
    <seriesInfo name="Internet-Draft" value="draft-geng-sidrops-aspa-analysis-03"/>
    <author initials="N." surname="Geng" fullname="Nan Geng">
      <organization>Huawei</organization>
      <address>
        <postal>
          <city>Beijing</city>
          <country>China</country>
        </postal>
        <email>gengnan@huawei.com</email>
      </address>
    </author>
    <author initials="M." surname="Huang" fullname="Mingqing Huang">
      <organization>Zhongguancun Lab</organization>
      <address>
        <postal>
          <city>Beijing</city>
          <country>China</country>
        </postal>
        <email>huangmq@mail.zgclab.edu.cn</email>
      </address>
    </author>
    <author initials="Y." surname="Wang" fullname="Yangyang Wang">
      <organization>Tsinghua University</organization>
      <address>
        <postal>
          <city>Beijing</city>
          <country>China</country>
        </postal>
        <email>wangyy@cernet.edu.cn</email>
      </address>
    </author>
    <date year="2025" month="August" day="25"/>
    <area>ops</area>
    <workgroup>sidrops</workgroup>
    <keyword>Internet-Draft</keyword>
    <abstract>
      <?line 69?>
<t>Autonomous System Provider Authorization (ASPA) is very helpful in detecting and mitigating route leaks (valley-free violations) and a majority of forged-origin hijacks. This document does an analysis on ASPA-based AS_PATH verification to help people understand its strengths and deficiencies, and some potential directions of enhancing ASPA are provided.</t>
    </abstract>
  </front>
  <middle>
    <?line 72?>

<section anchor="introduction">
      <name>Introduction</name>
      <t>Autonomous System Provider Authorization (ASPA) is a technique for verifying AS_PATHs in BGP updates <xref target="I-D.ietf-sidrops-aspa-verification"/><xref target="I-D.ietf-sidrops-aspa-profile"/>. Each AS can register ASPA records (also ASPA objects) in the RPKI to authorize a set of ASes as its providers. An AS can obtain ASes' ASPA records through RTRv2 protocol <xref target="I-D.ietf-sidrops-8210bis"/> and conduct AS_PATH verification based the records. ASPA-based AS_PATH verification can detect and mitigate route leaks violating the valley-free principle and path manipulations such as forged-origin or forged-path-segment attacks.</t>
      <t>ASPA can significantly enhance AS_PATH verification and is promising to be widely deployed. Despite of the strengths of ASPA, there are also some deficiencies. This document provides a detailed analysis on the strengths and deficiencies of ASPA. The document can help people deploy ASPA properly and provide some potential directions of enhancing ASPA.</t>
      <section anchor="terminology">
        <name>Terminology</name>
        <t>The usage of terms follows <xref target="I-D.ietf-sidrops-aspa-verification"/><xref target="I-D.ietf-sidrops-aspa-profile"/>.</t>
      </section>
      <section anchor="requirements-language">
        <name>Requirements Language</name>
        <t>The key words "<bcp14>MUST</bcp14>", "<bcp14>MUST NOT</bcp14>", "<bcp14>REQUIRED</bcp14>", "<bcp14>SHALL</bcp14>", "<bcp14>SHALL
NOT</bcp14>", "<bcp14>SHOULD</bcp14>", "<bcp14>SHOULD NOT</bcp14>", "<bcp14>RECOMMENDED</bcp14>", "<bcp14>NOT RECOMMENDED</bcp14>",
"<bcp14>MAY</bcp14>", and "<bcp14>OPTIONAL</bcp14>" in this document are to be interpreted as
described in BCP 14 <xref target="RFC2119"/> <xref target="RFC8174"/> when, and only when, they
appear in all capitals, as shown here.</t>
        <?line -18?>

</section>
    </section>
    <section anchor="aspa-strengths-and-disclaimers">
      <name>ASPA Strengths and Disclaimers</name>
      <t>ASPA records can be registered by an AS to authorize all its provider ASes. For the two ASes with mutual transit relationship, the two ASes will put each other AS into its own ASPA record (i.e., each AS considers the other AS as its provider).</t>
      <section anchor="protecting-against-route-leak">
        <name>Protecting Against Route Leak</name>
        <t>In full ASPA deployment (within a given region of interest), all "route leaks" (valley-free violations <xref target="RFC7908"/>) are detectable. Route leaks involve one of the following four valley-free violations:</t>
        <ul spacing="normal">
          <li>
            <t>A route is propagated through a P2C (Provider-to-Customer) link and then a C2P (Customer-to-Provider) link.</t>
          </li>
          <li>
            <t>A route is propagated through a P2P (Peer-to-Peer) link and then a C2P link.</t>
          </li>
          <li>
            <t>A route is propagated through a P2P link and then a P2P link.</t>
          </li>
          <li>
            <t>A route is propagated through a P2C link and then a P2P link.</t>
          </li>
        </ul>
        <t>It is expected that in partial ASPA deployment, not all route leaks are detectable.</t>
      </section>
      <section anchor="protecting-against-path-manipulation">
        <name>Protecting Against Path Manipulation</name>
        <t>Path manipulation can be path forgery or path tampering (i.e., insertion or removal of unique ASN) in this document. Forged-origin hijack and fake link-based hijack are all path manipulations.</t>
        <t>In full ASPA deployment (within a given region of interest), ASPA protects against a majority of forged-origin hijacks. Each AS can attest its upstream ASes, so provider or lateral peer cannot be deceived. Customer could be deceived because ASPA does not provide attestations to downstream ASes or peering ASes.</t>
        <t>Even in full ASPA deployment, not all path manipulation attacks can be detected. ASPA does not guarantee path correctness like that provided by BGPsec <xref target="RFC8205"/>.</t>
      </section>
    </section>
    <section anchor="aspa-deficiencies">
      <name>ASPA Deficiencies</name>
      <t>This section describes the deficiencies of ASPA-based AS_PATH verification in detail.</t>
      <section anchor="hard-to-detect-bogus-records">
        <name>Hard to Detect Bogus Records</name>
        <t>An AS can unilaterally authorize a set of its provider ASes. Under the one-direction authorization, an AS may intentionally or unintentionally register bogus records that are hard to be discovered. An AS maliciously registers bogus records that open a door to potential attacks.</t>
        <t><xref target="fig-bogus"/> shows an example of path manipulation attack based on bogus ASPA records. AS5 lies in that the nonadjacent AS3 is its provider in the ASPA record. The attack cannot be detected even when AS1, AS2, AS3, AS4, and AS6 register ASPA records correctly and enable ASPA-based AS_PATH verification locally. As a result, AS6 will wrongly consider its traffic to AS1 traverses AS5 and AS3, while the real forwarding path to AS1 is through AS4 and AS2.</t>
        <figure anchor="fig-bogus">
          <name>Path manipulation based on bogus ASPA records</name>
          <artwork><![CDATA[
              +-------+
              |AS1(P1)|Originate route
              +-------+
     path[1] /(C2P)   \
      +-------+        \(C2P)
      |  AS2  |         +-------+
      +-------+         |  AS3  |
 path[2,1]|(C2P)        +-------+
          |                 *
      +-------+            *
      |  AS4  |           *
      +-------+          *(faked C2P in AS4's
            \           * bogus ASPA)
  path[4,2,1]\(C2P)    *
              +-------+ ASPA{AS4, [AS2, AS3]}
              |  AS5  | Offender
              +-------+
      path[5,3,1] |(C2P)
              +-------+
              |  AS6  | Victim
              +-------+
  * All ASes register ASPA records
  * AS5's record states a faked C2P link with AS3
  * AS6 enables ASPA-based path verification
]]></artwork>
        </figure>
      </section>
      <section anchor="fail-to-detect-aspath-manipulation-by-a-provider">
        <name>Fail to Detect AS_PATH Manipulation by a Provider</name>
        <t>ASPA-based AS_PATH verification cannot effectively detect the AS_PATH maliciously shortened by a provider, which has been acknowledged in <xref target="I-D.ietf-sidrops-aspa-verification"/>.</t>
        <t><xref target="fig-path-shortened"/> shows an example. AS1 originates the BGP route and propagates the route to other ASes. The AS_PATH received by AS5 is path[4,3,2,1]. However, AS5 maliciously shortens the path by falsely claim a fake link with AS2 before AS5 propagates the route to AS6. AS6's traffic to AS1 may be hijacked by AS5 if the path[5,2,1] is shorter than any other AS_PATHs. In the example, AS5 may not intend to drop data traffic from AS6. That is, AS5 (provider) wants AS6 (customer) to prefer AS5's transit path for increasing revenue.</t>
        <t>In this case, the attack cannot be detected even when all the ASes register the correct ASPA records and all the ASes other than AS5 enable ASPA-based AS_PATH verification.</t>
        <figure anchor="fig-path-shortened">
          <name>AS_PATH maliciously shortened by a provider</name>
          <artwork><![CDATA[
                  +-------+
                  |  AS4  |    path[4,3,2,1]
                  +-------+----------------
      path[3,2,1]/         \               \
                /(C2P)      \               \(C2P)
        +-------+            \path[4,3,2,1] +-------+
        |  AS3  |             \             |  AS7  |
        +-------+              \            +-------+
  path[2,1] |             (C2P) \             /
      (C2P) |                    \ Offender  /
        +-------+ (Fake link) +-------+     /
        |  AS2  |*************|  AS5  |    /
        +-------+             +-------+   /path[7,4,3,2,1]
    path[1] |          path[5,2,1]|      /
      (C2P) |     (path shoterned)|(C2P)/(C2P)
        +-------+             +-------+/
        |AS1(P1)|             |  AS6  | Victim
        +-------+             +-------+
      Originate route
]]></artwork>
        </figure>
        <t>AS_PATH manipulation by a provider may also be used to do malicious route leaks. ASPA is not designed for defensing path manipulation. So, some malicious route leaks with path manipulation involved cannot be prevented.</t>
        <t><xref target="fig-malicious-leak"/> shows an example. AS2 is AS1's provider and arguably it may not leak its customer's prefix (P2) intentionally. But to increase revenue, AS2 may maliciously leak P2 with a modified AS_PATH (i.e., AS3 is removed) to AS4 for attracting more traffic to traverse AS2. Sometimes this attack may happen and goes undetected.</t>
        <figure anchor="fig-malicious-leak">
          <name>Malicious route leak</name>
          <artwork><![CDATA[
        +-------+
        |  AS4  |-------------
        +-------+             \
P1 path[2,1]|                  \
P2 path[2,1]|                   \ P2 path[3,1]
       (C2P)|                    \(C2P)
        +-------+ P2 path[3,1]+-------+
Offender|  AS2  |-------------|  AS3  |
        +-------+    (P2P)    +-------+
               \             /
      P1 path[1]\           /P2 path[1]
                 \(C2P)    /(C2P)
                  +---------+
                  |   AS1   |Originate route
                  | (P1&P2) |
                  +---------+
]]></artwork>
        </figure>
        <!-- ## Cannot Distinguish Leak and Hijack for an Invalid AS_PATH
Existing ASPA verification algorithm can identify Invalid AS_PATHs, but it cannot distinguish leak and hijack for an Invalid AS_PATH. The main reason is that ASPA records only focus on registering all provider ASes while not indicating the adjacency/topology information. When the Hop-check(x, y) function returns "Not provider+", the algorithm does not know whether the real cause of the result is i) ASy is ASx's customer/peer instead of provider or ii) link(x,y) is a fake link. Therefore, when the algorithm returns Invalid, it is not able to indicate whether the path is caused by a fake link-based hijack or not. 

For operators, it's instructive to know the type of an Invalid AS_PATH. If there exists no hijack, the Invalid AS_PATH is likely to be an unintentional route leak. Otherwise, the network may be attacked by path manipulation attacks.  -->

</section>
      <section anchor="not-directly-applicable-to-ibgp-ingress-and-ebgp-egress-verification">
        <name>Not Directly Applicable to IBGP Ingress and EBGP Egress Verification</name>
        <t>IBGP ingress verification and eBGP egress verification are meaningful in many scenarios. IBGP ingress verification is to check the AS_PATH received through iBGP connections. IBGP ingress verification helps an AS do verification on any BGP routers like non-ASBRs. EBGP egress verification means verifying the AS_PATH before sending it to the neighbor AS. It can prevent an AS from sending routes with Invalid AS_PATH to its neighbor ASes (just like eBGP egress RPKI-ROV <xref target="RFC8893"/>).</t>
        <t>However, current ASPA document <xref target="I-D.ietf-sidrops-aspa-verification"/> does not specify how to do iBGP ingress verification. For iBGP ingress verification, the router (e.g., an RR) conducting the verification may not have BGP sessions with the neighbor AS that propagates the route and thus does not know the local BGP role with respect to the neighbor AS. Even so, iBGP ingress verification is doable because the router can obtain the local BGP role from the ASPA records of local AS. In <xref target="fig-egress-veri"/>, when RR wants to do iBGP ingress verification, it can look up AS2's own ASPA records. If AS1 is listed as a provider, then apply the Downstream verification algorithm. If AS1 is not listed as a provider, then apply the Upstream verfication algorithm. Such an iBGP ingress verification also works correctly in RS and RS-client scenarios.</t>
        <figure anchor="fig-egress-veri">
          <name>IBGP and eBGP verification</name>
          <artwork><![CDATA[
          +-------------------------------+
          | AS2                           |
   path[1]| +-----+    +-----+    +-----+ |path[2,1]
AS1-------|->ASBR1|----> RR  |---->ASBR2|-|----->AS3
         /| +-----+   /+-----+    +-----+ |\
        / |          /                    | \
       /  +---------/---------------------+  \
      /            /                          \
  eBGP ingress   iBGP ingress              eBGP egress
]]></artwork>
        </figure>
        <t><xref target="I-D.ietf-sidrops-aspa-verification"/> does not specify how to do eBGP egress verification either. To try to do this, the router should add its own AS (i.e., AS2) to the AS_PATH and then performs ASPA-based AS_PATH verification from the perspective of next-hop AS (see Section 7.2 in version-15 of <xref target="I-D.ietf-sidrops-aspa-verification"/>). According to the verification result, the router decides to propagate the route or not. 
In <xref target="fig-egress-veri"/>, ASBR2 would add its own AS (i.e., AS2) to the AS_PATH. If the BGP role of ASBR2 with respect to AS3 is customer/lateral peer/RS/RS-client, the Upstream verification algorithm will be conducted. If the BGP role of ASBR2 with respect to AS3 is provider, the Downstream verification algorithm will be performed. The verification process also works well in mutual transit scenarios.</t>
        <!-- The problem of eBGP egress verification described above is that not all Invalid AS_PATHs (leaked routes) can be prevented from being propagated. Suppose the next-hop AS (i.e., neighbor AS3) is a lateral peer or provider. When the preceding AS (i.e., neighbor AS1) is also a lateral peer or provider but does not have an ASPA registered, the verification result can be Unknown (rather than Invalid). The route leak cannot be prevented in the case.  -->

<t>The relationship between AS1 and AS2 can sometimes be obtained by ASBR2 from AS2's ASPA records. If AS1 is listed as a provider in the AS2's ASPA records, then the above route leak can be detected and prevented. If AS1 is not listed in the AS2's ASPA records, ASBR2 cannot decide AS1 is a lateral peer or a customer. Therefore, the above route leak cannot be detected directly.</t>
        <t>Overall, iBGP ingress verification is doable with help of local AS's own ASPA records, while it is not possible to do eBGP egress verification correctly without more complexity.</t>
      </section>
      <section anchor="not-applicable-to-complex-relationship-scenarios">
        <name>Not Applicable to Complex Relationship Scenarios</name>
        <t>AS relationships in practical networks may be more complex than the traditional P2C/P2P model <xref target="as-rela-1"/><xref target="as-rela-2"/>. In ASPA, only the complex relationship of mutual transit relationship has been considered. The followings are some other complex scenarios that are not covered by ASPA:</t>
        <ul spacing="normal">
          <li>
            <t>Hybrid relationship <xref target="as-rela-1"/><xref target="as-rela-2"/>. Two ASes may agree to have different traditional relationships for different Points-of-Presence (PoPs). A hybrid relationship may be dependent on IP versions or/and PoP locations (see <xref target="fig-hybrid-rela"/> for examples), even prefixes (i.e., different relationships/policies for different prefixes).</t>
          </li>
        </ul>
        <figure anchor="fig-hybrid-rela">
          <name>Hybrid relationship</name>
          <artwork><![CDATA[
+-------+  Europe P2C  +-------+
|       |-------------->       |
|  AS1  |              |  AS2  |
|       |-------------->       |
+-------+   Asia P2P   +-------+

      (a) Location-dependent


+-------+   IPv6 P2C   +-------+
|       |-------------->       |
|  AS1  |              |  AS2  |
|       |-------------->       |
+-------+   IPv4 P2P   +-------+

      (b) IP version-dependent
]]></artwork>
        </figure>
        <ul spacing="normal">
          <li>
            <t>Partial transit relationship <xref target="as-rela-1"/><xref target="as-rela-2"/>. For a customer, the provider offers transit only toward the provider's peers and customers (or specific regions), but not the provider's providers, or restricts transit to a specific geographic region. <xref target="fig-hybrid-rela"/> shows an example. AS2 is the partial transit provider of AS1. AS2 should only propagate AS1's route to AS4 (i.e., AS2's peer) and AS5 (i.e., AS2's customer), but should not propagate the route to AS3 (i.e., AS2's provider).</t>
          </li>
        </ul>
        <figure anchor="fig-partial-transit">
          <name>Partial transit relationship</name>
          <artwork><![CDATA[
        +-------+
        |  AS3  |
        +-------+
  path[2,1] |
(route leak)|
            |
       (C2P)|
        +-------+  path[2,1]  +-------+
Offender|  AS2  |-------------|  AS4  |
        +-------+    (P2P)    +-------+
     Partial|    \                |
     transit|     \ path[2,1]     | path[4,2,1]
            |      ----------     |
    path[1] |           (C2P)\    |(C2P)
        +-------+             +-------+
        |  AS1  |             |  AS5  |
        +-------+             +-------+
      Originate route
]]></artwork>
        </figure>
        <ul spacing="normal">
          <li>
            <t>Persistent valley-path (legitimate valley-path) (<xref target="valley-path"/>). There may be legitimate valley-paths, i.e., violating the valley-free principle but the AS_PATH is legitimate. According to BGP data analysis, the persistent valley-path can be 10% of all BGP paths.</t>
          </li>
        </ul>
        <figure anchor="fig-valley-path">
          <name>Persistent valley-path</name>
          <artwork><![CDATA[
  Originate route
    +-------+             +-------+
    |AS1(P1)|             |  AS3  |
    +-------+             +-------+
            \             /
      path[1]\           / path[2,1] (not route leak)
              \(C2P)    /(C2P)
              +---------+
              |   AS2   |
              +---------+
]]></artwork>
        </figure>
        <t>ASPA records do not support the registration of complex relationships except the mutual transit relationship. As a result, in the complex scenarios, AS_PATH cannot be effectively protected by ASPA-based AS_PATH verification.</t>
      </section>
      <section anchor="reduced-protection-capability-in-partial-deployment">
        <name>Reduced Protection Capability in Partial Deployment</name>
        <t>To verfify an AS_PATH, ASPA verification algorithms need to check each hop of the AS_PATH. When ASPA records of the ASes along the path are partially registered, not all hops in the path can be checked. In such partial deployment scenarios, ASPA may have a reduced protection capacity.</t>
        <t><xref target="fig-partial-deploy"/> shows two examples of partial deployment. In <xref target="fig-partial-deploy"/> (a), AS3 cannot detect the route leak of P1 induced by AS2 if AS1 has no ASPA record registered. This is because the Hop-check(AS1, AS2) function returns "No Attestation" and the final verification result is Unknown. In <xref target="fig-partial-deploy"/> (b), AS3 is deceived by AS2 who falsely claims AS1 is AS2's neighbor. The attack in the example is undetectable because AS1 registers no ASPA record and AS3 cannot judge the validity of the link between AS1 and AS2.</t>
        <figure anchor="fig-partial-deploy">
          <name>Partial deployment</name>
          <artwork><![CDATA[
                            +-------+
                            |  AS3  |
                            +-------+
                            /
                           / path[2,1] (route leak)
   No ASPA                /(C2P)
  +-------+  (P2P)  +-------+
  |AS1(P1)|---------|  AS2  | Offender
  +-------+ path[1] +-------+
Originate route

      (a) Route leak in partial deployment

                            +-------+
                            |  AS3  |
                            +-------+
                            /
                           / path[2,1] (path manipulation)
   No ASPA (no adjacency) /(C2P)
  +-------+         +-------+
  |AS1(P1)|*********|  AS2  | Offender
  +-------+ path[1] +-------+
Originate route

      (b) Forged-origin attack in partial deployment
]]></artwork>
        </figure>
        <!-- ........................................................................ -->

</section>
    </section>
    <section anchor="reasons-of-aspa-deficiencies">
      <name>Reasons of ASPA Deficiencies</name>
      <t>This section summarizes three main reasons that result in deficiencies of ASPA:</t>
      <ul spacing="normal">
        <li>
          <t>ASPA record is authorized in one direction, e.g., an AS can unilaterally claim that another AS is its provider without the consent of other ASes. Related deficiencies:
          </t>
          <ul spacing="normal">
            <li>
              <t>Hard to detect bogus records</t>
            </li>
          </ul>
        </li>
        <li>
          <t>An ASPA record only focuses on including all provider ASes while ignoring other topology or relationship information. Related deficiencies:
          </t>
          <ul spacing="normal">
            <li>
              <t>Hard to detect bogus records</t>
            </li>
            <li>
              <t>Fail to detect AS_PATH maliciously shortened by a provider</t>
            </li>
            <li>
              <t>Not directly applicable to iBGP Ingress and eBGP egress verification</t>
            </li>
            <li>
              <t>Not applicable to complex relationship scenarios</t>
            </li>
            <li>
              <t>Reduced protection capacity in partial deployment</t>
            </li>
          </ul>
        </li>
        <li>
          <t>The kind of method that verifies AS_PATH based on relationships does not guarantee path correctness like that provided by BGPsec.
          </t>
          <ul spacing="normal">
            <li>
              <t>Not all malicious route leaks and path manipulations are prevented</t>
            </li>
          </ul>
        </li>
      </ul>
    </section>
    <section anchor="sec-security">
      <name>Security Considerations</name>
      <t>This document does not involve security problems.</t>
    </section>
    <section anchor="sec-iana">
      <name>IANA Considerations</name>
      <t>No IANA requirement.</t>
    </section>
    <section anchor="acknowledgements">
      <name>Acknowledgements</name>
      <t>Much thanks for the comments and inputs from Kotikalapudi Sriram.</t>
    </section>
  </middle>
  <back>
    <references anchor="sec-combined-references">
      <name>References</name>
      <references anchor="sec-normative-references">
        <name>Normative References</name>
        <reference anchor="I-D.ietf-sidrops-aspa-verification" target="https://datatracker.ietf.org/doc/html/draft-ietf-sidrops-aspa-verification-22" xml:base="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.ietf-sidrops-aspa-verification.xml">
          <front>
            <title>BGP AS_PATH Verification Based on Autonomous System Provider Authorization (ASPA) Objects</title>
            <author fullname="Alexander Azimov" initials="A." surname="Azimov">
              <organization>Yandex</organization>
            </author>
            <author fullname="Eugene Bogomazov" initials="E." surname="Bogomazov">
              <organization>Qrator Labs</organization>
            </author>
            <author fullname="Randy Bush" initials="R." surname="Bush">
              <organization>Internet Initiative Japan &amp; Arrcus, Inc.</organization>
            </author>
            <author fullname="Keyur Patel" initials="K." surname="Patel">
              <organization>Arrcus</organization>
            </author>
            <author fullname="Job Snijders" initials="J." surname="Snijders"/>
            <author fullname="Kotikalapudi Sriram" initials="K." surname="Sriram">
              <organization>USA National Institute of Standards and Technology</organization>
            </author>
            <date day="23" month="March" year="2025"/>
            <abstract>
              <t>This document describes procedures that make use of Autonomous System Provider Authorization (ASPA) objects in the Resource Public Key Infrastructure (RPKI) to verify the Border Gateway Protocol (BGP) AS_PATH attribute of advertised routes. This AS_PATH verification enhances routing security by adding means to detect and mitigate route leaks and AS_PATH manipulations.</t>
            </abstract>
          </front>
          <seriesInfo name="Internet-Draft" value="draft-ietf-sidrops-aspa-verification-22"/>
        </reference>
        <reference anchor="I-D.ietf-sidrops-aspa-profile" target="https://datatracker.ietf.org/doc/html/draft-ietf-sidrops-aspa-profile-20" xml:base="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.ietf-sidrops-aspa-profile.xml">
          <front>
            <title>A Profile for Autonomous System Provider Authorization</title>
            <author fullname="Alexander Azimov" initials="A." surname="Azimov">
              <organization>Yandex</organization>
            </author>
            <author fullname="Eugene Uskov" initials="E." surname="Uskov">
              <organization>JetLend</organization>
            </author>
            <author fullname="Randy Bush" initials="R." surname="Bush">
              <organization>Internet Initiative Japan</organization>
            </author>
            <author fullname="Job Snijders" initials="J." surname="Snijders"/>
            <author fullname="Russ Housley" initials="R." surname="Housley">
              <organization>Vigil Security, LLC</organization>
            </author>
            <author fullname="Ben Maddison" initials="B." surname="Maddison">
              <organization>Workonline</organization>
            </author>
            <date day="18" month="August" year="2025"/>
            <abstract>
              <t>This document defines a Cryptographic Message Syntax (CMS) protected content type for Autonomous System Provider Authorization (ASPA) objects for use with the Resource Public Key Infrastructure (RPKI). An ASPA is a digitally signed object through which the issuer (the holder of an Autonomous System identifier), can authorize one or more other Autonomous Systems (ASes) as its upstream providers. When validated, an ASPA's eContent can be used for detection and mitigation of route leaks.</t>
            </abstract>
          </front>
          <seriesInfo name="Internet-Draft" value="draft-ietf-sidrops-aspa-profile-20"/>
        </reference>
        <reference anchor="RFC7908" target="https://www.rfc-editor.org/info/rfc7908" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.7908.xml">
          <front>
            <title>Problem Definition and Classification of BGP Route Leaks</title>
            <author fullname="K. Sriram" initials="K." surname="Sriram"/>
            <author fullname="D. Montgomery" initials="D." surname="Montgomery"/>
            <author fullname="D. McPherson" initials="D." surname="McPherson"/>
            <author fullname="E. Osterweil" initials="E." surname="Osterweil"/>
            <author fullname="B. Dickson" initials="B." surname="Dickson"/>
            <date month="June" year="2016"/>
            <abstract>
              <t>A systemic vulnerability of the Border Gateway Protocol routing system, known as "route leaks", has received significant attention in recent years. Frequent incidents that result in significant disruptions to Internet routing are labeled route leaks, but to date a common definition of the term has been lacking. This document provides a working definition of route leaks while keeping in mind the real occurrences that have received significant attention. Further, this document attempts to enumerate (though not exhaustively) different types of route leaks based on observed events on the Internet. The aim is to provide a taxonomy that covers several forms of route leaks that have been observed and are of concern to the Internet user community as well as the network operator community.</t>
            </abstract>
          </front>
          <seriesInfo name="RFC" value="7908"/>
          <seriesInfo name="DOI" value="10.17487/RFC7908"/>
        </reference>
        <reference anchor="RFC2119" target="https://www.rfc-editor.org/info/rfc2119" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.2119.xml">
          <front>
            <title>Key words for use in RFCs to Indicate Requirement Levels</title>
            <author fullname="S. Bradner" initials="S." surname="Bradner"/>
            <date month="March" year="1997"/>
            <abstract>
              <t>In many standards track documents several words are used to signify the requirements in the specification. These words are often capitalized. This document defines these words as they should be interpreted in IETF documents. This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements.</t>
            </abstract>
          </front>
          <seriesInfo name="BCP" value="14"/>
          <seriesInfo name="RFC" value="2119"/>
          <seriesInfo name="DOI" value="10.17487/RFC2119"/>
        </reference>
        <reference anchor="RFC8174" target="https://www.rfc-editor.org/info/rfc8174" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8174.xml">
          <front>
            <title>Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words</title>
            <author fullname="B. Leiba" initials="B." surname="Leiba"/>
            <date month="May" year="2017"/>
            <abstract>
              <t>RFC 2119 specifies common key words that may be used in protocol specifications. This document aims to reduce the ambiguity by clarifying that only UPPERCASE usage of the key words have the defined special meanings.</t>
            </abstract>
          </front>
          <seriesInfo name="BCP" value="14"/>
          <seriesInfo name="RFC" value="8174"/>
          <seriesInfo name="DOI" value="10.17487/RFC8174"/>
        </reference>
      </references>
      <references anchor="sec-informative-references">
        <name>Informative References</name>
        <reference anchor="I-D.ietf-sidrops-8210bis" target="https://datatracker.ietf.org/doc/html/draft-ietf-sidrops-8210bis-21" xml:base="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.ietf-sidrops-8210bis.xml">
          <front>
            <title>The Resource Public Key Infrastructure (RPKI) to Router Protocol, Version 2</title>
            <author fullname="Randy Bush" initials="R." surname="Bush">
              <organization>Arrcus, DRL, &amp; IIJ Research</organization>
            </author>
            <author fullname="Rob Austein" initials="R." surname="Austein">
              <organization>Dragon Research Labs</organization>
            </author>
            <date day="11" month="June" year="2025"/>
            <abstract>
              <t>In order to validate the origin Autonomous Systems (ASes) and Autonomous System relationships behind BGP announcements, routers need a simple but reliable mechanism to receive Resource Public Key Infrastructure (RFC6480) prefix origin data, Router Keys, and ASPA data from a trusted cache. This document describes a protocol to deliver them. This document describes version 2 of the RPKI-Router protocol. [RFC6810] describes version 0, and [RFC8210] describes version 1. This document is compatible with both.</t>
            </abstract>
          </front>
          <seriesInfo name="Internet-Draft" value="draft-ietf-sidrops-8210bis-21"/>
        </reference>
        <reference anchor="RFC8205" target="https://www.rfc-editor.org/info/rfc8205" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8205.xml">
          <front>
            <title>BGPsec Protocol Specification</title>
            <author fullname="M. Lepinski" initials="M." role="editor" surname="Lepinski"/>
            <author fullname="K. Sriram" initials="K." role="editor" surname="Sriram"/>
            <date month="September" year="2017"/>
            <abstract>
              <t>This document describes BGPsec, an extension to the Border Gateway Protocol (BGP) that provides security for the path of Autonomous Systems (ASes) through which a BGP UPDATE message passes. BGPsec is implemented via an optional non-transitive BGP path attribute that carries digital signatures produced by each AS that propagates the UPDATE message. The digital signatures provide confidence that every AS on the path of ASes listed in the UPDATE message has explicitly authorized the advertisement of the route.</t>
            </abstract>
          </front>
          <seriesInfo name="RFC" value="8205"/>
          <seriesInfo name="DOI" value="10.17487/RFC8205"/>
        </reference>
        <reference anchor="RFC8893" target="https://www.rfc-editor.org/info/rfc8893" xml:base="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8893.xml">
          <front>
            <title>Resource Public Key Infrastructure (RPKI) Origin Validation for BGP Export</title>
            <author fullname="R. Bush" initials="R." surname="Bush"/>
            <author fullname="R. Volk" initials="R." surname="Volk"/>
            <author fullname="J. Heitz" initials="J." surname="Heitz"/>
            <date month="September" year="2020"/>
            <abstract>
              <t>A BGP speaker may perform Resource Public Key Infrastructure (RPKI) origin validation not only on routes received from BGP neighbors and routes that are redistributed from other routing protocols, but also on routes it sends to BGP neighbors. For egress policy, it is important that the classification use the 'effective origin AS' of the processed route, which may specifically be altered by the commonly available knobs, such as removing private ASes, confederation handling, and other modifications of the origin AS. This document updates RFC 6811.</t>
            </abstract>
          </front>
          <seriesInfo name="RFC" value="8893"/>
          <seriesInfo name="DOI" value="10.17487/RFC8893"/>
        </reference>
        <reference anchor="as-rela-1" target="https://www.usenix.org/system/files/nsdi19-jin.pdf">
          <front>
            <title>Stable and Practical AS Relationship Inference with ProbLink</title>
            <author>
              <organization/>
            </author>
            <date year="2019" month="February"/>
          </front>
        </reference>
        <reference anchor="as-rela-2">
          <front>
            <title>Inferring Internet AS Relationships Based on BGP Routing Policies</title>
            <author>
              <organization/>
            </author>
            <date year="2011" month="January"/>
          </front>
        </reference>
        <reference anchor="valley-path" target="https://ieeexplore.ieee.org/document/6363987">
          <front>
            <title>Valley-free violation in Internet routing - Analysis based on BGP Community data</title>
            <author>
              <organization/>
            </author>
            <date year="2012" month="November"/>
          </front>
        </reference>
      </references>
    </references>
    <?line 388?>



  </back>
  <!-- ##markdown-source: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-->

</rfc>
