Monday, February 13, 2006
NGN and Customer Value
Last week I participated in the ETSI TISPAN Plenary #10, taking place in Sophia Antipolis near Antibes in France. After celebrating the (in)completion of TISPAN NGN Release 1 end of last year, TISPAN is heading now under the new chairman Rainer Muench full steam for Release 2.
So what can be expected from TISPAN NGN Release 2? The operators made this quite clear: Tele Danmark, British Telecom, TeleSonera, Telecom Italia, Swisscom, Telenor, France Telecom and Deutsche Telekom presented a joint paper (10TD218) proposing an initial list of high priority items for TISPAN NGN Release 2:
The major points are: user equipment, PSTN/ISDN simulation, support of IN services (INAP, CAMEL), 3G Profile A, SIP-I Profiles B and C, simulation services (CW, ACR, AoC, CCBS, MCID, ...), localized and global mobility management, IPTV and VoD, PtT, SMS, MMS, support for corporate networks, IP PBX, Centrex, Overload and Congestion Control, Admission Control in RACS to handle QoS over core, expand IP-CAN for other access networks than xDSL (3G RAN, WLAN, Cable networks), define procedures for roaming and handover, NGN roaming for WLAN scenarios, IPv6, NAT traversal, Presence, IM, UPSF/HSS data, PSTN/ISDN Emulation, QoS, Policing and Congestion Control, etc. etc. etc
I counted approx. 60 (sixty) such high priority items. Don't ask me about the low priority items.
I cannot imagine half of this to be done within two years, and then the question still exists if all this stuff will work and interwork.
Since I was reading at the same time Geoff Hustons article about Convergence?, I was asking if TISPAN is considering what the customer wants. No customers are present in standard bodies. The operators say that they are of course taking the customer needs also into account. How well they are doing this shows history: e.g. ISDN, IN, ATM (and IMS?). So the basic question is: is a standards body capable at all to define services? Or should this be left to the application providers?
The problem here is that one should not implement what the customer wants, but what he values = what he wants to pay for. Best example: a customer wants QoS, but he is not willing to pay for QoS, at least not if he gets reasonable QoS 99% of the time with best effort. And one cannot live on the 1% overflow traffic.
It is not quite true that there where no customers present at TISPAN: there was ECMA, explaining their need for the NGCN (Next Generation Corporate Networks). Their presentation (10TD190) was quite telling: of course it was basically a letter to Santa Claus, but in essence what they wanted was to be connected to the NGN of the operators, but only as a last resort. They want of course to interconnect within their corporations and also between corporations directly, via leased lines or via the Internet. Rich Shockey showed in his presentations already that this is about 80% of the traffic. If the NGN now can make a living out of this 20% overflow?
There is also another problem lurking nobody is yet really looking at: the NGNs are also marketing their network as more secure (e.g. no spam, no viruses - Note: yesterday I saw a message that 50% of MMS messages contain viruses - so much for security). This implies that the NGN service is connected to the CPE or corporate LAN via a VPN. But this also implies that there is NO connection to the applications running at the LAN (e.g. Outlook or corporate databases) and also NO connection to the Internet. Since a normal end-user and also most companies are connected to the Internet, you either bridge this at the customers LAN (and then you have access from the Internet into the NGN) or you prevent this, then there is no interworking possible.
This is NOT what customers want.
So what can be expected from TISPAN NGN Release 2? The operators made this quite clear: Tele Danmark, British Telecom, TeleSonera, Telecom Italia, Swisscom, Telenor, France Telecom and Deutsche Telekom presented a joint paper (10TD218) proposing an initial list of high priority items for TISPAN NGN Release 2:
- items postponed from Release 1
- capabilities to enhance Release 1
- new network capabilities in support of new services and applications
The major points are: user equipment, PSTN/ISDN simulation, support of IN services (INAP, CAMEL), 3G Profile A, SIP-I Profiles B and C, simulation services (CW, ACR, AoC, CCBS, MCID, ...), localized and global mobility management, IPTV and VoD, PtT, SMS, MMS, support for corporate networks, IP PBX, Centrex, Overload and Congestion Control, Admission Control in RACS to handle QoS over core, expand IP-CAN for other access networks than xDSL (3G RAN, WLAN, Cable networks), define procedures for roaming and handover, NGN roaming for WLAN scenarios, IPv6, NAT traversal, Presence, IM, UPSF/HSS data, PSTN/ISDN Emulation, QoS, Policing and Congestion Control, etc. etc. etc
I counted approx. 60 (sixty) such high priority items. Don't ask me about the low priority items.
I cannot imagine half of this to be done within two years, and then the question still exists if all this stuff will work and interwork.
Since I was reading at the same time Geoff Hustons article about Convergence?, I was asking if TISPAN is considering what the customer wants. No customers are present in standard bodies. The operators say that they are of course taking the customer needs also into account. How well they are doing this shows history: e.g. ISDN, IN, ATM (and IMS?). So the basic question is: is a standards body capable at all to define services? Or should this be left to the application providers?
The problem here is that one should not implement what the customer wants, but what he values = what he wants to pay for. Best example: a customer wants QoS, but he is not willing to pay for QoS, at least not if he gets reasonable QoS 99% of the time with best effort. And one cannot live on the 1% overflow traffic.
It is not quite true that there where no customers present at TISPAN: there was ECMA, explaining their need for the NGCN (Next Generation Corporate Networks). Their presentation (10TD190) was quite telling: of course it was basically a letter to Santa Claus, but in essence what they wanted was to be connected to the NGN of the operators, but only as a last resort. They want of course to interconnect within their corporations and also between corporations directly, via leased lines or via the Internet. Rich Shockey showed in his presentations already that this is about 80% of the traffic. If the NGN now can make a living out of this 20% overflow?
There is also another problem lurking nobody is yet really looking at: the NGNs are also marketing their network as more secure (e.g. no spam, no viruses - Note: yesterday I saw a message that 50% of MMS messages contain viruses - so much for security). This implies that the NGN service is connected to the CPE or corporate LAN via a VPN. But this also implies that there is NO connection to the applications running at the LAN (e.g. Outlook or corporate databases) and also NO connection to the Internet. Since a normal end-user and also most companies are connected to the Internet, you either bridge this at the customers LAN (and then you have access from the Internet into the NGN) or you prevent this, then there is no interworking possible.
This is NOT what customers want.
Comments:
Post a Comment
