Internet Engineering Task Force 103, 6 Nov 2018

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: IETF103-IPWAVE-20181106-1120

Description

IPWAVE meeting session at IETF103
2018/11/06 1120

https://datatracker.ietf.org/meeting/103/proceedings/

A

B

B

Erik's looking for you.

B

Okay, can we find a seat so we can get started please.

B

The chair, stop. First, with the chairs have some things to say.

B

Okay, as usual, the first thing is blue sheets. We have one clipboard going around, please make sure you sign it second thing is: we need a jabber scribe and a note-taker.

C

B

Have one note-taker, okay, need a jabber scribe.

C

Okay, so let's start so, as you saw the intellectual property that you must be aware,.

C

And the note well.

C

Okay, so the blue sheets are in circulating, we have note-taker and Russ will be in Joppa.

C

So this is the agenda for today's session. We'll do some initiative stuff at the beginning and then we'll focus into them into working with documents, the one about ipv6 over dot, 11 ocv and then the Primus teaming and use cases document and then time allowing we will have a short presentation and neighbor discovery for ipv6 and another one on TLS using 80s. But this is time permitting so on.

C

We have one a slide on the liaison.

B

Yes, we got a liaison statement from ITU SG 16, basically informing us of some work they're doing they.

B

This was posted to the mail list a while ago. We received it on the 3rd of September. They just want to make us aware of this work and if we have anything to say about it, they'd love for to hear about it, but basically they're calling for volunteers to participate in this focus group. If you're interested you don't need to come through us, just talk to the ITU directly focus groups, don't have the same formal membership requirements as some other ITU activities.

B

So if you look at that liaison statement and you're interested go ahead, thank you. I think you're right.

C

So short working with status update, so we have you know two documents in working group process, the one the first one isn't: the transmission of ipv6 over all the yield 2.11 of seaweed. This document has been stuck for a while due to some six-month trace issues. Concerns will see just after this presentation from the authors and then we will discuss how to move forward on this document, and then we have another document.

C

That is the promise teaming and use cases document that has been through some reviews, but we still need some some additional reviews to to make this document move to the next stage. We will discuss that later as well, so these the status that you can see we are late on on the original on the original deadlines. We are very much aware of that. We will we'll see how to actually speed up things during the meeting.

C

So, on the next steps we will, as I mentioned, discuss the the review has been done on the ocv document and see whether the human race issues still remain.

C

We believe that there are things that are still to be done and we are discussing some some ways of approaching this or addressing this, maybe adding some new litter may may help, but is still under discussion, and but we believe that document is not ready yet and we will need to do some work and then submit to the exhibit, but not before the six-month basically agrees with the document. As stated in our charter that the six-month has to review the document that we produced.

C

These are the first one on the second one and the survey Pro and statement use cases document. We got some preview since last ITF million- that's very good because we managed to get something. That is a good starting point, as actually some other reviewers summation, but we need. We do need additional reviews.

C

I mean this is starting point, but we need more mature reviews on this iteration, not only to improve the quality of the document to a level that is good enough to for publication, but also to to provide some proof that the working room has energy to do additional stuff, because otherwise there is no point in having a praana statement document. If there is no problem, then to be solve, so we knew we really need at least two additional detail reviews.

C

So it's not about typos editorial kind of review, but really on the on the on the content of the document before ITF 104, and we need a new version of the document before the meeting and by before the meeting. I don't mean two days before the submission deadline, but at least one full iteration before so. Basically, two additional versions before idea, 104, otherwise I guess we have been waiting for quite a long time, so we will be may be forced to drop the document.

C

If there is no energy on this, we will discuss that later as well.

C

That's basically it from from our side. So, let's move to the next point in the agenda: Nabeel.

D

Hi, this is Nabila banana from Morocco I'm Anna I soak returning fellow to ITF, so I'm presenting our draft transmission of ipv6 packets offer. I triple e a so 2.11 networks operating in mode outside the context of basics and resets. So.

D

Since the last ITF meeting that was held in Moria, the drafts is updated with all recent comments, since that ITF, including comments from six men working group from Eric and from the human rights protocol, consideration that helped us to consider all this statement and to include it in the current version of the draft.

D

So as part of efforts of the human right motor consideration. Amelia has reviewed the human rights consideration and suggested to remove the reference to that RFC, which is 880 280, because it doesn't need to be listed in the reference as it's an IRC F research group of documents and its review is meant to be helpful and in no case a mandatory.

D

So the main comments that we have received from 6-man are related to impart to the neighbor discovery. So the main neighbor discovery challenge were suggested by six men and by mr. eric, not Marc. They were addressed in version 26 and 30. 30 is the latest version that we have. The respective change are listed in the changelog.

D

So if you want to have more says about all the chances that we have between version 26 and 30, you can just compare the chance rocks, so they are in for in in 30 in the version 30 a clarification on the reliability of ND /, o CB + / 8 result of 11 links. So we added the following paragraph: the neighbor discovery protocol is used over eight or two dots, 11, o CB links. The reality of the ND protocol over 80, so totally philosophy is the reliability of the delivery of nd multicast messages.

D

This reliability is the same as the rarity of delivery of nd multicast messages over 80 total TV links operated with bssid. So we have at the data paragraph and we are waiting for comment from Eric. So ok comment from from 6 miles, so in so for the 26th version. So we move to text so.

E

If you go back to the site, so the Eric nutbag, so the second paragraph is a sorry second sentence is actually incorrect. Two things is incorrect. One is Andy uses both unicast and multicast. 2 nd actually sends multiple messages. So if a single message, it's lost its you're, not dead right. So the wording here can need some use of improvements.

E

The thing that's important to keep in mind for OCD is that unicast is as unreliable as multicast, which is different than Wi-Fi right. What people have done on Wi-Fi is saying not ok if you can switch to unicast or better off, because it's more reliable we have ads, but but it might make sense to state that fact that OCB, because it doesn't have acts because there are associations, even justices, badass multicast or is good whatever that's fine, but but you should actually fix that second sentence, because I think it confute in you to confuse people.

E

D

Thank you very so you suggest to rewrite the second sentence in this paragraph yeah.

E

D

Focus also on unicast.

E

Yes- and you could actually weave it in I- could suggest some words, but okay.

D

Thank you. Thank you for your comments.

D

Back to version 26 of the draft, so we moved the text from slack section and from design consideration append this about privacy into a new privacy. Consideration subsection of the security section, as has been suggested by in in the mailing lists, or if M, rated, the slack and ID section to stress only link locals, does, can use UI, 64 removed the GOP, Wireshark explanation or reformulated the slack and ll sections.

D

Continuing on the comment from six men, we added a brief mention of the need of use of link local clarified to text about mark address changes, dropped the Dhoni discussion and change title of section describing an example of packet forma for the privacy protections of vehicles, which is, it has been also comments that we have received from deck and Emilia. So we added that paragraph that you can see in the so we need you to add this paragraph in future.

D

Release of the of the document.

D

So it's it's basically about the demand of for privacy, protection of vehicles and drivers identifies which could be granted by using episode, name or areas. Identity at the same time may hamper the required confidentiality of message and trust between participants.

D

So that's that's all for deserves. Thank you.

C

Okay think reveal so I guess you didn't get like kind of the blessing from six-month right. I mean the last revision has not been reviewed by six month as such. So we need that, because I I believe that the sixth month will not be happy with the current version either. So we need to to probably get another review from them to highlight the things that need to be addressed.

F

So I one more thing is like probably also the check with the IPV working group to see if they're happy with the changes, because I don't see like you know, like the closing the loop in the working group. So maybe we can probably do like a one-week last call or something not this week, but the week after say like okay, like you know, are you happy with the changes that are being made since the last book in the last call, because I think it's pretty significant amount of changes, since the last working group last call?

F

So once that's done, probably I'll do the same process for six month to ship it over, but I. Think doing this in lockstep is probably not gonna work, so we can probably close one off and then go again do another shot, because I think it's gonna be very difficult to try to do a last call on both the places at the same time, because there might be somebody who might come up like even like stuff.

F

Like you know, Eric was talking about like some fairly obvious things that, like shouldn't, be controversial, seem to be controversial right, like some appendix or something right, so I I, just don't want to like you know, correlate these two things at the same time. So like kill it close off your send it off the six man, bring up the issues here again and gone. So even if it takes time, I think it makes more sense and that way.

E

Eric again so two things, one I was looking through the six man comments and there's at least one that hasn't been addressed, which is the paragraph after the nd paragraph that refers back to that paragraph. As the motivation for you know, mobile.

G

E

Whatever and people pointed on I didn't know when this force in the last couple months, but that that back reference doesn't make sense, given that you now changed the paragraph it's referring to, so that's something that needs to be addressed, I think in general, it works well when there's email responses, saying okay, based on what you said, we've changed this section to this, whether it's six months comments to what I made I have to go.

E

Look at the diffs to actually figure out what was actually changed right because to change log in the document is extremely brief, so I haven't pulled up the diffs yet to see what actually did change but but doing that stuff so that people can actually see that. Okay, you did change something in response to this, or at least you responded saying: oh no you're confused. You know we don't need to change the document. Okay,.

D

F

That that's kind of what I was implying by closing the loop right, like you know, there's like people, don't know that you made the change like our. What change actually tried to make for a given things like at least four Eric's email right, like you listed like you know, what are the changes he made, but all the six man comments right doesn't there's no closure in six-man for the comments that came in right, saying, okay, this is what we changed in the graph.

F

This is why we're not making a change in a draft and so on for the issues that came up there right.

D

As far as I know, so we have received many comments from six men and we addressed all the comments because it was everything.

F

D

F

The media list with.

D

Carpentry and Lorenzo.

G

Lorenzo exactly.

D

So we have answered in the million list you can, you can see in the middle, so there are some point that we didn't agree, so they were not impacted in the in the draft, but we didn't get other questions from them or comments from them. Okay, okay,.

F

So we just probably go through it like separate, be useful for you, Carlos Andres. If I set up an issue tracker for this, so you can actually keep track of the issues and close them off and and go on, because I feel it's very weird like to get an issue tracker here to go through this like side process to get one. But I can get your issue tracker, for this is that you think I'll have I'm.

B

Not sure the list is so long, we need a issue tracker. We just compile. One email says this was your issue. This is what how we address them. Don't you think? That's enough! That's fine! Okay, okay, perfect.

C

So then, just to summarize, the process will be I. Can there are some comments pending for being updated in the drugs right? Do.

G

C

Guys meet a new version with all the changes that this thing at this point were necessary. You send an email summarizing all the changes you made were those comments came from like six month. It says it is this a wave. We have address a comment. If it is a comment, didn't reach an agreement, you say like we didn't reach an agreement, so this is still open.

C

Do summarize everything once we have that we go for working group last call one week in the working group and after that, what whatever comments we get and a new version potentially addressing that we asked reg to go to six month again right, perfect thanks.

D

Thank you very much. Thank you.

H

Hello, this is a war zone from sang-gyun University in Korea. The bottom of the slide lists up our co-authors. So at least talking about about the problem statement addressing what is next, the staff for IP wave. If we can and also show the several use cases, the actor material, meaning we updated the two of virgins so the especially as a comment from volunteer divers Eric and the Turk and prong choice and be sure and Ally yeah: okay, Amalia, okay, so except the Amelia. That is a human lie to leave you.

H

So what is lighted at the human? Try to point of view so highlight are her the lib you most of focus and the security stop. So I tried to address especially privacy issue. So I can polish up our problem statement document, but other technical part I, believe I addressed all so we agreed just a Picasa 3. Our key work items that the first one is neighbor discovery. The second one is the mobility management. The third one is security and privacy.

H

So the major of changes is a neighbor discovery. I would say, pick your neighbor discovery if we Andy. So last time the adding node mark, he asked the what is definition of a particular link model. So this drug we tried to address the link model, so maybe we can discuss this very important. The second one is the neighbor discovery based proactive, a handover to support the transporter layer session continuty. So we can take advantage of Lincoln layer parameter such as: let's see the channel power indicator and also location of our ICU and it's a coverage.

H

So we can predict when the mobile vicar will move out over the communication. Ranger RCU and the Odyssey also can determine what is the timing to perform the handover. So we can take advantage of that information, so that is a second one. Our sort of one is a vicar law on that architecture. Try to clarify the Cure link model, the last one is also our see you on that. Talk is connected to a vehicle cloud, so I will explain one step at a time. This figure shows the picture to Architecture.

H

So you can see the bottom. We have two 7s 7 at 107, a 2, so this figure is very important, so you can see the RSU you 1rs you to to others. You are sharing the same prefix, so we try to construct the multi-link a subnet, which means one beaker p1 configured by the prefix announced by RSU one and then he moving to another R to you. Well, it's a virtual long link.

H

It doesn't necessarily be configure our game because the same muscle in cassava net, we tried to reduce address of the computation overhead right because you can see usually our ICU or will be deployed intersection earlier, sometimes highway cases of intermediate over our bottle. Aways. Most of the cases, the urban area, the particular people think intersection, is a good position, so you can see the usually intersection. Distance is less than 100 meter in the United States case. In the case, every RG you visit changing new address is not efficient, so I envision.

H

We try to construct a more technical subnet and then we can allow just a one-time others configuration. It will be enough. However. Sometimes another RC you three can have another estaba net. Prefix, so in that case, it's somehow we needed to perform some handover, so 3 and I get discussed so mobility anchor. This structure, similar to a proximal arrive here, such as RMA. So this one can hinder our different subnet, prefix I, think that is it some different from proximal variety structure.

H

So we try to take advantage of the position of original and trajectory of a vicar because the vicar can equip with GPS navigator. So we can lock on our easy positions, feed the direction that information really adequately deported to RSU, and you can share that information with the neighbor Ric you also. It can be poor to mobility anchor, so mobility anchor is a kind of control, caravan or mobility management. So I think this is a new link of model.

H

Also important thing is, secondly, the big helicase, the communication between picker to beaker shield p2p later than a pre, I and I to be it takes time. So we sport, b2b or so pre I simultaneously. This one is I think a different from traditional or infrastructure face the wall, its communication, so we need to support two posts, so this figure shows up with two I communication scenario you can see. This is the left hand, side the picker in peak collateral and the right hand.

H

Sides are see you in internal network, so nowadays the industry they are trying to make internal lateral, using Gigabit Ethernet, providing some diagnosis or other some sensing for some adaptive cruise control purposes. Many there are can be generated, so we can see this one is moving router. Okay, it is a communicate with another router.

H

This is the keep in mind the RSU, so this interface is used to access cloud connected to a poverty anchor to take a some internet services over communicate with this vicar. Another vicar locate he to add another southern, a blink okay. So we needed to consider this kind of internal letter as well. Okay, our second one is that the beaker to bicker case can communicate each other. So in this case we can say one important: the application in the driverless autonomous our vicar scenario case. We can learn some the control context-aware navigator.

H

They can communicate to each other to aware the neighboring vicars also neighboring are pedestrians to avoid conflict or accident. So in the case of some application, learning some hosted to another cruise control application laying another host within in beaker Network. They should communicate to each other. So the question is how to provide this kind of communication, so that is a different from other existing manner or other some DMA and other Asian area.

H

Okay, so I try to clarify the link model are based on mainly from the adding, not mat, so the legacy ipv6 a neighbor discovery protocol is not suitable for PQ loyalists link because the major concern is a symmetry in connectivity.

H

You can see the VQ okay depending on transmission power, also interference in air, so one transmission not necessarily guarantee the other, a transmission right, so a symmetry in the connected exist and also the. Lastly, you can see the multi-link a serve on a case. Take xu the leech over, even though the multi-link a so connect consists of a multiple or wall.

H

It's linked with the same prefix, so the how to hinder, and also the peculiar case, the initial attack Network case, even though sometimes within some subnet domain, but maybe take disconnected because of a some communication range right. So in the case we need to communicate using our seu relay something like that. So so this is very different from traditional.

H

The IP version, 6 neighbor discovery protocol assumptions, so we needed to create qualify.

H

So as a leader we have some new features for particular neighbor discovery, so Eric suggested are lightweight. Dat are using a ND optimization from 6lowpan 6lowpan. The major concern is energy consumption, so they try to avoid nd a broadcast. So if some IOT device requires some address prefix case, just unicast based router, solicitation message delivered and louder can be saved and then handler using multi hub at the ad.

H

So in the same way we take advantage of this one imagine you can see the one intersection area 100 over 200 beakers moving around what they try to reconfigure. So in that case we need to minimize traffic overhead. That is I, think important. Also it moving to another subnet. Try, the idea again is, you can say, inefficient approach. So that's why we need to consider this one. So single address, comparation, more technical, subnet I think is enough.

H

Also the handle bar among our seu and the MA mobility anchor can be used. Support, take advantage of our trajectory and the position information over the baker. Also, we have their dynamic, topology ARPANET, also multi-link a subnet forwarding. We need consider P P over P IIT bleah, so we new consider that one depending on the location of Vickers, okay and then, let's move on our mac address pseudonym for privacy.

H

So in the case back address a pretty radically change effects, address change, IP layer, so mini hinder, okay, Sally, yeah.

I

My name is Charlie Perkins. Can you go to the last slide just for a moment? Okay, there's a document, that's just about done in six loads called sixty authorities, 775, update and I- think it really deserves consideration here, because there's a lot of discussion about how to have multi-hop registration for these devices as well, as you might say, some improved verification capability which might come in handy here as well. Okay, so.

H

B

Is it the name hooking up with the.

I

Rfc number: it's not an RFC, yet it's in the RFC editors queue but I mean they're, they're finishing it, but it's RFC 6775 update as a solo.

H

Document: okay,.

I

H

Okay, thank you. Yeah.

E

Just to clarify it's an update to that RFC, so it has some additional pieces associated with it. So but it's fun it's the same model right. It's just adding something: okay,.

H

Good yeah, thank you very much, so the MAC address pseudonym case.

H

The important thing is current. My idea, the MAC addresses shoot on in case the change in IP address compact, the TCP such.

G

As the transport.

H

Layer session should be the maintained right, so we need to consider all that, so we can take advantage of existing our mobility management scheme. Maybe it is not good enough. We needed to adding more and also v2. Visionary. Okay is prefixed dissemination exchange very important, I explain the previous slide free to be scenario and the mobility management, so efficient mobility management, simulates connectivity and timely data exchange.

H

So we can take advantage of a mobility information sharing or with the infrastructure load, our ICU and mobile T anchor and the peakers so proactive a handover can be used. So I delivered the trajectory page. The mobility management can be used. Okay, nowadays, every user or driver holding a smartphone over thicker nowadays a desk the system providing the vision system. We can take advantage of that mobility information for our mobility management.

H

Also with that kind of prediction over the vicar mode model T, we came for from efficient the da de terre poverty and the handover for active manner. There is item very important, unique point and also we take advantage of a link of layer parameter such as receive the channel power indicator over our SSI over the outer layer frame. We cannot use it so next step is Carol mentioned that we text to intensive live year.

H

Maybe the three: the conduct belly right, three and other expert, maybe the volunteer about the mobility and some Mary Eric leave you again, maybe so we can move out and then we need to reach other and they be moving out to develop some protocol. Okay, thank you. Okay,.

C

So for the the reviews, so we need to three reviews before next ITF, so you suggested three I, don't know if you can do it, okay, so one thank you provides, certainly what's also mesh. Well, I, don't know if mentioned I was thinking in some it tell me. Can you take a look at the at the current version of provides and comments as well.

J

Thank you and any other volunteer. I'm Rahul Jadhav.

K

From our technologies, I have a question regarding the mobility scenarios that I've mentioned. One of the things that wave has done is removed. Some handshakes and they've updated the authentication scheme at the lower layer, just so that it reduces additional additional are td's. You know that I know more and now at the ball layer with 60s 67 75 updates. We are going to introduce.

K

So what I was wondering is: is there any information in the draft which which gives this information that you know this is the hard bound that is kept by the way of standard? Oh, what is okay.

H

First ISO: your concern is about some security issue. No, it's not security, so the.

K

Layer to view has changed their handshake mechanism so as to reduce the number of parties that are involved for certain scenarios. Right and now we are adding those are tt's at the bow layer. So what I'm trying to understand is? Is there any hardbound that has been placed by the wave standard on the on what what amount of time can be spent in this? This authentication, handshake or any other handshake that is involved at the I'm, just worried that at the hive layer we are introducing those handshake and increase in the time bound.

K

You know, okay,.

G

K

It is there any specific considerations that you have done for that particular aspect. Okay,.

H

Those currently as I long as understand the wave case say they are focused on three to be scenario lighter than three to I, so because the handover usually involved with the IP right so communicate with some other corresponding located to some other subnet. So, but that is exactly our IP way will help or working on that. So I think that is a no reference. Yet as long as I understand, okay yeah. Probably that is a very important yeah. You right.

K

So I read the draft last time. I didn't see such information, but then I tried to check whether such any any any such hard bound is placed by the way.you standards. I didn't find anything specific, but I just wanted to know whether you.

L

Know or something like.

H

K

Okay, thank you.

L

So Rob oh you're, saying that is more able to set up the time. Okay. So, but here we are not talking about the solution approach, yet I I know even Paul is talking about. Even the is talking about the different mobility management approaches. We are not sure if you're going to build it layer to the me. You know, one approach is RS here, routed or a bridge all right.

L

Essentially in one model, you can have some higher-level aggregation K to build a large l2 domain right like today, how capped at this right, if you keep it very close to kpop kind of architecture. That is one we need to look at how that impacts. Our you know different mobility management scenarios, but if you look at RSU as more as a l3 router, then what Paul is talking about like a handoffs and all of that in that model we need to say how it impacts, so we haven't discussed any of the solutions.

L

So this is just a high-level comment at this point. This is because it's a problem statement document. These are all potential problems, but yep I think you need to look at 16 or fix what exactly is a shoe, but okay.

H

Thank you, the three. Yes, okay, sorry, hello,.

M

My name is Sammy and I understood that I wrote the human rights review that you put to be decided, and so I would just like to have clarified for myself, which parts of the review that were difficult to understand or mm. Why the concerns were into are.

H

You the Amelia right, yes,.

M

H

Is very helpful is false. Only the private issue of in number vehicle identification number issue, other synonym I think a very humble I tried to address a some apart, but some apart, it is looks like they're not easy to yeah, so we can discuss okay after some.

G

H

Okay, thank you.

E

H

E

Not necessarily comment on this graph, but it might sort of fall in between the two graphs and so I realized going through the various comments that well. One thing that we haven't specified as far as I know is the base case of the vehicle is talking to one RSU and the second later is talking to another one, because it drove down the road right and how does that actually perform today and what sort of level of tuning is needed? You know we have these various or a advertisement interval.

E

We have ways of setting router lifetimes and reachable time and other things right and I. Don't know if anybody's done any experiment without with sort of current code and see how does this behavior if a new RSU shows up every second or whatever right I mean even if he you you're, not trying to get global routing to work, but you just want to be able to ping the default router as you're driving it on the road right.

E

That would be useful test case to see how quickly can you actually discover that there's a new default router, you shouldn't use the old one because you lost connectivity right and.

H

I, don't think it requires protocol.

E

Changes, but at least you need to figure out what sort of tuning is needed here. So that's a lot simpler than this. Your your mobility management here assumes that you will find out about I'm now talking the new RSU I should stop talking to do block right and I can run my mobility protocols on top of that, but you need to find that out in a recently timely manner, so mm-hm anybody on experiment.

H

For this, so actually you know the completest region about the white IP wavy case, the other eleven o CB driver and the you know system is not popularly available. So, for example, on my lab purchased Kota Wallace RC, you are very expensive, so it's fair amount taken either so I think the.

H

Alternatively, my lab are working in simulation based, realistic, later-model, fading model, try to peer out what is the limitation of the current yet existent or so testify next slide, we tried to come up with a new some approach for Lincoln model and the neighbor discovery mobility issues, so we are working on assimilation, so hopefully, next the problem eating I bring my student to try to demonstrate what is the limitation away this team protocol right? It's the our proposal, so at least the simulation after that we can moving out there. Implementation.

E

Yeah I mean we can talk offline about ways possible to run experiments to if okay, thank.

C

You yeah, okay, thank you, so any additional volunteer for reviewing the document. As we discussed, we need to three reviews before we even consider going for working group last cause we got to, although one is from Charlie that is I just realized, that is a contributor coaster, but I mean anyway. That would be helpful. The other one is for three and any other volunteer I.

H

Recommended John: he also he's a support yourself: okay, okay,.

C

So I take it offline, but you need to really find at least two.

G

More reviewers.

C

Okay, okay, then we move to next.

H

Okay, this one is try to come up with a some possible candidate for I 2016 neighbor discovery for IP waste bicular network, so the Jeong jeong's, Shang and Christian are my student.

H

So the introduction so motivation of the vehicle, our neighbor discovery is. We need to some concrete to some idea. So I would say this is not necessarily a solution, but it can provide some kind of line for our father moving places. So the subjects are handled by this. A trapped is definition of the link model for vehicle over wireless links now. Secondly, neighbor discovery, optimization with multi-hop tid and the soda was proactive handover with pixel unable discovery in the mobility management. Last one is our MAC addresses. You don't need Henry.

H

So again, this is a picture. You understand the right so so we try to provide b2b and the P to I. Consider our merger linkers of Annette scenario so again, UK the vicar is a registered through our ECU and this address and actually register into a mobility anchor. So this link coverage case the its neighbor castes entry having the vicar's address information. Also that address information is delivered to our mobility anchor.

H

So as a result, another vehicle try to register using this, our c2 and our to try to register address here if address come completely happen, which means the daily fared. Okay, it is called the multi half deity and then notify, and then it tried modified, configure again otherwise register. So in this mechanism, we can provide that one shut. Tid is enough. Okay and then it can move in out so mobility point of view we can using the proper IP it can handle okay, okay, three so.

L

When you say register, is it like a proxy? What exactly is our do.

H

Kind of proxy, so let me show you this one, this one, so the beaker or sending other try to send an S with the rest, restoration and then honestly received so first of all, it has a neighbor cast, it can cache it each address and then using duplicate the rest restoration. This one has a 30, ATK /, the rester take off look at its the ad table. It pretty conflict. The case this to address confirmation, saying there is a complex.

H

Otherwise, okay is okay and then it can acknowledge, okay, the assignment and then definitely not okay. So.

L

So again, going back to my earlier comment: if you make look at our SC as an access point right and essentially, mobility anchors, like you know, somewhat an aggregation gateway like our Sun controller. Alright, if you look at kpop architecture now, if do you still need this kind of bridge? You know so if now in that l2 model right, how is it impacting this call flow here? In other words, you know keep everything all the state on a mobility anchor make RSU more as l2 pipe right between our SEO and mobility.

L

Anchor is just layer to pipe now. How does this impact your scenario here so.

H

Your point is the mobile anchor is a kind of the some. The powder-like point try through using the TMS approach.

L

All I'm saying is rsu's just say: l2 l2, Oh.

H

So currently yeah RSU, my assumption is rs3 is the l3 as well so.

F

H

Some this is some justice which case probably some louder, the behind that the our ICU, but the might have finishes RC, is at the router, see.

L

If you deploy RSC every 300 meters right now, if I'm going at 90, km/h I think this model may not work I think we need to really really think whether RSU is an IP node or at l2 node and, like I, think we need to look at the scenario. Maybe you know, building a large l2 domain is one approach, I think, but but we need more discussion. Yeah you're.

G

L

You know understanding how it is solving the problem here, but with respect to all the scenarios, but what we need to look at yeah.

H

Okay, so okay go ahead. Address yeah.

F

Suresh krisshnan so, first to like make sure everybody understands. This is like a future proposed work item for this working group right. We really want our deliverables done before anything like this happens, and the second thing that I found confusing I don't know if other people did is the relationship between this draft and brass young right and they like kind of interrelated. But it's not like very clear how they're interrelated, so I would really recommend that you try to put them together so because I'm I don't really understand why they're separate oh.

H

Okay, so actually the Chinese are my draft and showing the my student. We in the picture try to merge at the moment that we take a refer to some options, but in the revision case that we can merge right so.

F

That's the thing because, like this one uses like option definition.

G

Or the other draft and they're a little back and.

F

Read it to make sure I understand.

G

F

I think it's causing some confusion so take care of that, but I would really like not discuss the details just give up like you know, we don't have too much time right so just use the time to, like you know, provide like an elevator pitch like two minutes of the thing and okay sure, okay, good yeah,.

H

So I think there is a main idea and that we can hinder pseudonym and the mobility take out the panty here. What are se? You I explained the problem statement, the document, so our next step is once we are finished a problem statement. We need to start never this copy for p2p with UI. That is my intention, this document, so to proof of concept we improvement.

H

Currently we are implemented line. Now we finished up the basic labor discovery protocol implementation. Next step, we implement our proposer. So next the hackathon project we try to come up with so as a one information I, also leading some I to another hackathon, we seven times, hackathon alternates, working, very energetic. We you can tackle it, the working guru, a Lara. You know what interrupt so more than sixty seventy percent are our ad groups, so I want bring in energy into this walking or using the deer walking code. Thank you for your comment. Okay,.

C

Okay, one question or comment regarding this.

H

C

Project, we are also a word that I think now. Will you have some implementation work going on right on the basic ipv6 overall CV with some students? That's correct or not. So my question is: if you do also, you have also some implementation on the basic ipv6 over ocv stuff, because if that is the case, it will be very valuable. You go for this kind of hackathon, where you do it some interrup tests on the basic draft.

C

That will be very helpful for that draft to to progress, to understand what are the issues and then that will be also very hopefully understanding. What are the potentiality that need to be improved in a potential future work, as shouldas mentioned, nothing will happen until we are really done with them into documents.

C

So I think we should focus on those two main documents so basically on the implementation of the ocv basic stuff, and that will be very very helpful if you can do a hackathon kind of interrupt in the ex idea and if that's the case, we should prepare and advertise. So many, maybe some other implementations or some other people can join. Ok, but it will make mainly meet the hackathon around the ipv6 oversee the main stuff. Ok,.

D

This is noble again. Thank you. Thank you for this suggestion we have indeed in the last IES african internet sandwich. We have done a hackathon and there were three trucks, one of them it was around a IP wave, so we brought some network cards rated to OCB and it was a successful because we have been able to implement those cards in in Linux and Linux recompile, the the current etc, and make the pink between two to two notes that there are virtual vehicles because of dot cards, and we got some good results. Yes, so.

H

Yeah I would write a kilometer with novels theme and my lab, and there we can take lovely to write and then we can move out what should be first. Maybe the neighbor is cover option. Ok, ok, thank you. Thank you. Thank you. So.

C

Let's move to the last one.

N

Hi I'm Mira from Kampala tech present to you today the the extension of transport layer, security code to support the vehicular certificate at sea and I Triple E certificate. So the agent of this presentation, I, will really start by these currents. Our motivation use cases of the extension and the extension how it works on terra's, so cooperative, intelligent transport system are very, very mobile, highly mobile system with a limited bandwidth. That's why we are not using knowing certificate unknown security tools.

N

That's why we are using size, optimized certificates standardized by each year and I Triple E Etsy is one out of three: oh nine, seven and I Triple E. Sixteen, oh nine point to the Etsy. One is just a profile of the I Triple E 1, so it has the name, the same data structure.

N

Our objective is to make this authentication and to define the topic authentication method, more optimize it for bandwidth and the processing time to set for delay sensitive application and to enable client and server authentication in cas network.

N

Some use cases for this extension is the upload of vehicle data and the vehicle. Let's say: log server to log, for example, to upload it from vehicle to log server and to update the vehicle software and the vehicle configuration also.

N

We can use it also for to connect to make connection to cloud services connected on containment and also to connect a client vehicle to to company using daily servers like rent company, rent company and car manufacturers, and also wireless electrical, the electric vehicle charging, the extension, how we make it really in Telus it just adding a new certificate type to existing types, x.509 or a public key and the third one. It will be 1609 the two and to define its structure.

N

Entry structure on on TLS and here I'm, describing an example of handshaking where the client search the client is proposing to use X 1609, the two and the server, and he gives to the server the possibility to use one of this authentication method. The first one x.509, the second one, is raw public key and the third one is 1609 the two, and so the plant will keep his certificate 1609 dot, two for authentication and the server will choose the default, one which is x.509. It's a hybrid utilization of such certificate with x.509.

L

Basic cushion so in dot two 69.2 by default there, how many certificates out there so there's one from the OEM break or there is it- can provide.

N

Some clarity on that it has the same it has as an industry as an inn data structure and inside it you can define which profile you want to, and Etsy is one profile on I Triple, E.

L

N

It's it. Let's say the general data structure is proposed by I, Triple, E and Etsy, and we can you can profile rust, you can define a profile and this general structure so.

L

Within that profile, if it's I Triple E profile, is that, like a manufacturer certificate or some other, it's.

N

It's I mean if so, it's a profile, so it's we verify the certificate in the same way and we serve its it's the same fields just for example, on Etsy. Some fields are optional, are deactivated by default and on I Triple E is more general, let's say definition, but on the verification you we have to follow the algorithm, the verification algorithm of I Triple E, even for HC certificate. Okay,.

M

N

So so here we we define a new value referring to the I Triple E certificate, to add it to the certificate type as a client certificate, type and server certificate type, as defined it on latest tailless FC. Thank you.

C

Very much so with this we are done for today. Thank you very much. Where are the blue sheets by the way.

C

Thank you. Thank you.

G