First sign of a future home :)

Raw walls now, but now you can believe it will be home one day :)

Posted by email from kgoworek's posterous

Gathered a team to do ALS on a dying project

12 people, 10 days to save a project that was initially estimated to take 200 mandays.

Posted by email from kgoworek's posterous

Sunday at work

I'm spending a lovely Sunday at work. It is also my birthday. :-(

Posted by email from kgoworek's posterous

Operators stop IMS development

From a recent article at telecoms.com:

At the [...] IMS Strategies conference held in Dusseldorf two weeks later, a major European operator declared it was effectively pulling out of IMS core development for the foreseeable future.

The operator that made this declaration is KPN.

News that KPN is concentrating on developing new services on its all-IP core - without the benefit of IMS for the foreseeable future at least - is not entirely bad either. Now the hype cycle is complete, it has become clear that investment in IMS is a long-term affair.

[...] As a means of responding to more immediate threats such as competition from, for example cablecos looking to become MSOs, IMS is worse than useless in the short term. This is most likely the reason why KPN has put the brakes on IMS development. It does however beg the question of why what is seen as one of the meanest, leanest operators in Europe didn't think these issues through before embarking on a well-publicized strategy of developing an IMS-based core network some 12 months ago.

[...] some smaller operators have already stated publicly that [...] IMS is not on their roadmaps for similar reasons.

The efficacy of horizontal service delivery architectures has already been proven to offer cost savings and reduce rollout times, however the same cannot be said for IMS and at present, its role in a single service environment remains unclear.

It seems that after several years of IMS hype the air is starting to get out of the balloon. One thing doesn't change, however. Operators do invest in all-IP networks, they still want Service Delivery Platforms.

It seems probable that no single technology / architecture will be accepted soon by most telcos, we'll live in a heterogenic world (I guess it's typical to the packet-switching era). What it means is more work and more money for system integrators.

Vendors cause IMS implementation hold-ups

In the recent issue of "Telecommunications" Iain Morris wrote that British incumbent BT blames network equipment vendors for hold-ups in IMS implementation.
There is no vendor that provides equipment 100% compatible with the IMS standard. Of course, all the vendors provide "almost-IMS" solutions that work but not interwork.

Malcolm Wardlaw, BT’s director of converged services says

"When vendor equipment meets the standard it will be a lot easier for carriers like BT to mix and match vendors, and that’s critical both commercially and from a resilience point of view,"

And another quote from the article:

According to a new report from consultancy Analysys, that problem could be endemic. "It is difficult to produce a business case for a generic capability like IMS without being specific about the services that it will support, and compelling service propositions have not yet been forthcoming,"

BT strategy is to provide wireless services through partnerships with mobile operators. Lack of IMS-compliance makes this task impossible.
It seems that the two parties (operators and vendors) have different views on the situation.
As we can see, telcos say that vendors are not ready for IMS. Vendors say that most network operators are not ready for IMS — they have no business cases and no well-defined migration strategies.
I'm afraid that almost nobody is really ready for IMS yet. So it is possible that another standard architecture becomes obsolete before it gets implemented.

Softswitch vs IMS

When you talk to telecom equipment vendors, they say that the era of softswitch technology ends, and softswitches will soon be replaces by mature IMS platforms.
However, instead of abandoning softswitch technology, telcos deploy more and more of them every year. Do they give up IMS instead?

First, in this post, let's say what softswitch and IMS are.

IMS technology is a further development of softswitch technology. Softswitch architecture separates service control from service access, and uses an IP-based core layer in the switching network.
All-IP IMS technology uses IP network to access terminals network. It also uses IP-based protocols (including SIP — Session Initiation Protocol — and Diameter) for signaling purposes.
Softswitch is a voice solution whereas IMS is designed for multimedia capabilities.

Softswitch is more or less a software-based version of traditional digital exchange. It provides IP telephony service. In IMS architecture, it is analog to CSCF (Call Control function). Softswitch implements service logic to control external trunking gateways, access gateways, and remote access servers. Softswitches run on commercial computers and operating systems, and they provide open applications programming interfaces. Also called a call agent or a media gateway controller. Softswitch cooperates with SS7 network. It can either work with PSTN phones or SIP phones.

IMS is a complete architecture that contains application servers, call control function, home subscriber servers, media gateways and servers etc. Softswitch can be incorporated into IMS during the process of network evolution.

Massive migration to all-IP networks?

Latest Heavy Reading report states that telcos are getting ready for the carrier network transformation — from CS / ISDN technology to all-over-IP.
It seems we are facing huge investments in IP network transport equipment. The thing that will probably go in parallel will be investments in new IMS service platforms. Most important reason of network migration seem to be the OPEX savings (softswitch/IMS solutions are cheaper to run and maintain than PSTN/ISDN exchanges) and new innovative applications and services.

Less and less people believe that a new killer-app will emerge to become a driver for the migration. Most of IMS services can be realized now, without migrating the whole network infrastructure. Some of them can be deployed on IN platforms, some — on IMS application platforms interoperating through media gateways with CS-based networks.
So, in my opinion, the migration will not happen so soon, at least in the emerging markets. New and/or small operators will for sure decide to build IP-based infrastructure but incumbents and large GSM operators are not so eager to invest so much so fast...