Re: [secdir] draft-zorn-radius-pkmv1-05.txt

"Glen Zorn" <> Wed, 26 August 2009 16:58 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 82A6D28C28B for <>; Wed, 26 Aug 2009 09:58:50 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -2.642
X-Spam-Status: No, score=-2.642 tagged_above=-999 required=5 tests=[AWL=1.957, BAYES_00=-2.599, GB_I_LETTER=-2]
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id 4Af63k0TZk7p for <>; Wed, 26 Aug 2009 09:58:49 -0700 (PDT)
Received: from ( []) by (Postfix) with SMTP id E0D3B3A68E9 for <>; Wed, 26 Aug 2009 09:58:48 -0700 (PDT)
Received: (qmail 7131 invoked from network); 26 Aug 2009 16:58:55 -0000
Received: from unknown ( by ( with ESMTP; 26 Aug 2009 16:58:54 -0000
From: Glen Zorn <>
To: 'Donald Eastlake' <>
References: <> <00b701ca24de$8f53e4a0$adfbade0$@net> <>
In-Reply-To: <>
Date: Wed, 26 Aug 2009 09:58:49 -0700
Organization: Network Zen
Message-ID: <001501ca266e$7bd0f3f0$7372dbd0$@net>
MIME-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
X-Mailer: Microsoft Office Outlook 12.0
Thread-Index: AcomUEf16wCeDaZiTCOBHwvBgKiHxwAEIXvw
Content-Language: en-us
x-cr-puzzleid: {29146CB4-ACC6-48CB-8CCD-A93F679ACD90}
Cc: 'Dan Romascanu' <>, 'IETF Discussion' <>,
Subject: Re: [secdir] draft-zorn-radius-pkmv1-05.txt
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Security Area Directorate <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Wed, 26 Aug 2009 16:58:50 -0000

Donald Eastlake [] writes: 


> >> This document defines seven RADIUS Attributes to support the
> >> implementation of 802.16 (WiMax) PKMv1 (Privacy Key Management
> version
> >> 1). I would guess that RADIUS can be used between the 802.16 Base
> >> Station and an authorization server but I don't know how you could
> >> tell.
> >> Maybe I missed it but it looks like the RADIUS protocol isn't
> >> mentioned anywhere in 802.16-2004. From the text in some of these
> >> RADIUS attribute descriptions, it appears that they are not used
> >> between the Subscriber Station and the Base Stations but may be the
> >> basis of 802.16 Attributes that are used on that hop. Given this, I
> >> think a paragraph is needed (maybe even accompanied by a little
> >> art) at the beginning show what's going on would be useful.
> >
> > Your comments seem to suggest a lack of familiarity with RFC 2865 and
> the
> > RADIUS protocol in general.  Leaving aside the question of how one
> could
> > expect to usefully review a document that _extends_ a protocol w/o
> > understanding the protocol being extended, RADIUS is only defined
> between a
> > NAS (in this case, an 802.16 Base Station) and a RADIUS server.
> You have missed my point entirely. To evaluate extensions to RADIUS
> whose purpose is to support 802.16 seems to me to require not just a
> knowledge of RAIDUS but some knowledge of how RADIUS fits into 802.16
> and what is needed for RADIUS, with these extensions, to support
> 802.16 security. My point was that, so far as I can tell, how RADIUS
> fits into 802.16 isn't documented in your draft or any document
> referenced by it. 

Right, that's because, as you have discovered & is the raison d'être of this
draft, RADIUS does not fit in with 802.16-2004 at all.  The relationship
between RADIUS and 802.16-2004 is basically identical to that between RADIUS
& PPP: RADIUS is a simple bolt-on allowing the centralization of
authentication & authorization data (& concomitant scaling & configuration
benefits).  You can search the set of RFCs defining PPP for a mention of
RADIUS, but that search would be in vain even though RADIUS has been (&
still is) used in virtually every PPP (PPPoE, PPPoA, ...) deployment.

> Doing a little more research, 802.16e-2005, which
> you don't reference, does begin to touch on this by at least
> specifying how EAP fits into 802.16.

I don't reference 802.16e-2005 because it's irrelevant to the draft; in
fact, it's an almost completely different protocol.  802.16-2004 defines
what is colloquially referred to as "fixed WiMAX", while 802.16e-2005 forms
the basis of "mobile WiMAX".  Mobile WiMAX has a lot more features, is
arguably more secure and is just generally sexier (mobility is all the rage,
you know ;-).  For all these reasons & undoubtedly more of which I am
unaware, the 802.16 WG has moved on & is no longer interested in fixed
WiMAX; that doesn't mean that it is not useful, though.  For example, the
public access network at the recent Olympic games in Beijing was based upon
802.11 using 802.16 backhaul to a pre-existing FDDI ring.  This made
networking basically the entire city a project of months rather than years,
since the nearly impossible task of running physical cables was unnecessary.
As it stands, fixed WiMAX is well suited for this kind of relatively static
application, since the appropriate credentials, authorization data, etc. can
be configured into the base stations and left pretty much alone.  However,
this kind of network design (.11<->.16<->some kind of wire) is also very
appropriate in more dynamic situations; for example, building access
networks in developing countries where the challenges & costs of running
physical cabling would be even more problematic.  Tack up a few .11/.16
gateways, plug in a .16 base station & voila!  Instant Internet.  So why not
just use mobile WiMAX & just not move (much ;-)?  The problem is that all
those sexy mobility features make designing, testing & manufacturing mobile
WiMAX equipment much more expensive (obviously a major consideration).  The
problem is that these instant networks are also quite dynamic & the base
stations themselves may not be conveniently located, so it would be nice to
not have to configure the BS every time somebody hangs a new .11/.16 gateway
on a tent pole or schoolhouse wall.  RADIUS support solves this problem for
fixed WiMAX in the same way it did for PPP.

> If above you are saying that the security of these new RADIUS
> attributes can be evaluated entirely based on a knowledge of RADIUS, I
> do not agree with this. 

Why not?

> If above, you are saying is that there is no
> need for there to be some explanation, in your draft or in some
> document referenced by it, of how RADIUS fits into 802.16 and that
> people who don't have an a priori knowledge of this should just keep
> their noses out of your document, I don't agree with that either. RFCs
> are supposed to be for the Internet community and there is presumably
> a reason that, for example, the RFC Editor requires acronyms that are
> not widely known to be expanded. Of course, it is not necessary or
> possible for most RFCs to be self contained and understandable in
> isolation, but I think it is reasonable to expect them to refer to
> other documents so that someone willing read enough and follow the
> reference chains can get a clear picture of what is going on.
> Thinking about it, I suggest that the Abstract be changed to something
> like the following and a similar change be made in the Introduction.
> (I note that the Introduction has a fairly long paragraph giving many
> details about 802.16 PKMV1 while failing to shed much light on how
> RADIUS or EAP fit into 802.16.)

As I mentioned above, that's because neither EAP nor RADIUS fit into
802.16-2004, which is the only standard that is relevant.

> "RADIUS can be used by an IEEE 802.16 Base Station, acting as an EAP
> Authenticator, to communicate with a remote Authentication Server to
> authenticate 802.16 Subscriber Stations and support 802.16 Privacy Key
> Management Version 1. This document defines a set of additional RADIUS
> Attributes which are designed to enable this support."
> If you make this change and, due to my inferior knowledge, I've made
> any errors in the above paragraph, I'm sure you can fix it.

Fine.  The thing is that, except for the "acting as an EAP Authenticator"
part (which is false), I can't really see how this imparts much more
information than the existing text (assuming familiarity w/RADIUS).  But if
it will make you happy...

> >> Many document have security considerations section that only refer
> to
> >> other documents and may be missing specifics to the document
> contents.
> >> I think this document has the opposite problem good security
> specifics
> >> in the security consideration section but could usefully add
> >> references to the 802.16-2004 and RADiUS security sections.
> >
> > I'm not at all sure what 802.16 security has to do with RADIUS, but I
> guess
> > I can add a reference to RFC 2869 in the Security Considerations
> section.
> Because you are extending RADIUS for the express purpose of supporting
> the 802.16 base station security function of authenticating subscriber
> stations. This is a base station function whose mechanism, as far as I
> can see, is glossed over in IEEE Standard 802.16-2004, which you
> reference but which has absolutely no mention RADIUS, NAS, AAA, EAP,
> Authentication Server, or anything along those lines. As above, by
> further research, I found that the IEEE 802.16e-2005 amendment to
> 802.1-2004 at least mentions EAP. I believe that 802.16e-2005 should
> be added to the references in the draft.

As I mentioned above, 802.16e-2005 is completely irrelevant to this draft.  

> Is it really such a burden, to provide some security context for what
> you are doing, to say something like:
> "Security consideration for IEEE 802.16 appear in Section 7 of
> 802.16-2004 and of 802.16e-2005. Security considerations for RADIUS
> extensions appear in RFC 2869." or the like?

No, it's not much of a burden for me.  OTOH, I've read section 7 of
802.16-2004 a bunch of times & really haven't found anything like the
Security Considerations section of an RFC.  Furthermore, there's a huge
amount of stuff (e.g., traffic encryption key management, etc.) in there
that is, again, irrelevant to the draft and potentially confusing to the
non-experts that I assume would be the intended audience.


> >> The wording in Sections 3.1 and 3.2 see to almost be designed to
> allow
> >> the possibility of the multiple *-Cert Attributes carrying a
> >> certificate to appear in more than one Access-Request message. But I
> >> would assume that's not meaningful and/or was not intended to allow
> >> that.
> >
> > There is no way to do such a thing in standard RADIUS.
> That's what I thought and why I was puzzled as to why there was a more
> complex wording that appears to permit this. I suppose it is just the
> way the words struck me at the time I read them. But I would, instead
> of
>           If multiple PKM-SS-Cert
>       Attributes are contained within an Access-Request packet, they
>       MUST be in order and MUST be consecutive attributes in the
> packet.
> have said
>       These multiple PKM-SS-Cert Attributes MUST appear consecutively
>       and in order within an Access-Request packet.
> and similarly for PKM-CA-Cert.


> Looking at this more closely now, I also think that the final "s"
> should be removed from "instances of the PKM-SS-Cert Attributes" and
> "instances of the PKM-CA-Cert Attributes".


> >> The table of attributes in Section 4 that gives the number of times
> >> each attribute can occur in different message types seems to have
> >> problems. Since there is no key giving it another meaning, I assume
> >> "0-1" means zero or one. But PKM-SS-Cert and PKM-CA-Cert are
> described
> >> as possibly occurring multiple times due to fragmentation of
> >> certificates. If the table is supposed to be in terms of logical
> >> attributes so that multiple PKM-SS-Cert attributes only count as one
> >> if they have parts of one certificate, then the table should say so.
> >> On the other hand, the PKM-SA-Descriptor attribute is shown as "0+",
> >> which presumably means zero or more, but the text description in 3.6
> >> clearly says it can occur one or more times, which presumably would
> be
> >> written "1+".
> >
> > The relevant text from section 3.6 says "One or more instances of the
> > PKM-SA-Descriptor Attribute MAY occur in an Access-Accept message."
>  RFC
> > 2119 says about the keyword "MAY", "This word, or the adjective
> > mean that an item is truly optional"; this says to me that zero, one
> or more
> > instances of the PKM-SA-Descriptor Attribute can be in an  Access-
> Accept
> > message, just in a more compact and formal way.  If this is not
> clear,
> > however, I'm open to suggestions for alternate text.
> The wording is OK but I would suggest making it one letter longer so
> it starts "Zero or more ...".

Isn't that redundant?