[AfrIPv6-Discuss] «Preparing for IPv6-only mobile networks : why and how»

[JORDI PALET MARTINEZ]

I was considering the background of the article: cellular networks with smartphones. I don’t think even the test mentioned in the article considers 3G/4G modems.

So in this case, you don’t run NAT444, just NAT44.

You don’t need to “manually” check the phone of each uses. It works creating a dual-stack APN and an IPv6-only APN, you can detect the cellular phones/OS version and configure to them one or the other. So users with IPv6-enabled phones (using 464XLAT such as Windows and Android, or with IPv6-enabled apps such as iOS), will use IPv6-only connectivity (but they get dual-stack service in the phone).

Smarpthones not capable of IPv6-only, or not IPv6-enabled at all, will use the dual-stack APN.

Regards,
Jordi


-----Mensaje original-----
De: Mark Tinka <mark.tinka at seacom.mu>
Responder a: <mark.tinka at seacom.mu>
Fecha: jueves, 2 de junio de 2016, 20:00
Para: Jordi Palet Martinez <jordi.palet at consulintel.es>, IPv6 in Africa Discussions <afripv6-discuss at afrinic.net>
Asunto: Re: [AfrIPv6-Discuss] «Preparing for IPv6-only mobile networks : why and how»

>
>
>On 2/Jun/16 16:39, JORDI PALET MARTINEZ wrote:
>
>> I don’t think NAT444 reduces cost even initially … You need to buy the boxes, you have more radio bandwidth used for the keep-alives, etc., even more power consumption.
>>
>> It may be a case by case, for example if you existing network doesn’t support IPv6, or the majority of your uses doen’t have smartphones with IPv6 support, etc.
>
>The amount of effort required in making sure your customer's phones
>support IPv6, as well as having the back-end systems updated to match,
>is too daunting for most mobile operators, taken from the point of view
>of what they can immediately gain, but spending a couple of millions of
>NAT444 today to earn billions tomorrow.
>
>The desire for change has to be fundamental.
>
>Mark.
>