Some DSL (Digital Subscriber Loop) providers have certainly led the industry in communicating rates for their service that are unlikely to ever be experienced by their customers, with statements such as "Speeds upto 50 Mbit/s,"  note that all-important 'upto' term.  Ofcom in the UK has started an investigation into this practice.  Physical loop limitations mean that unless your house is next to the local exchange, the actual throughput will be much lower, typically one half to one quarter of that rate.

Earlier this year I did a quick survey examining the mobile industry and its claims as multiple wireless technologies are offered to customers.  The US is focusing upon average download speeds and defining the conditions upon which those speeds are achieved.  The rest of the industry uses 'upto' with in some cases guidance on what can typically be expected.

For DSL, through-put is dependent on loop-length, noise at the local exchange (generally Gaussian), concentration in the carrier's network, and congestion on the site being accessed.  For example, at home I know I get 2.8 Mbit/s rain or shine.  However, for my mobile data service the throughput is dependent on my distance from the tower, the number of people simultaneously using that tower, whether I'm indoors, the weather (fading is Rayleigh not Gaussian, so sometimes it just does not work), which device I'm using, and what smarts the operator has in their network to squeeze the data; as well as the concentration in the carrier's network and congestion on the site/service being accessed.

My experience with throughput on my mobile service is definitely more variable.  This raises the question does average download speed mean anything?  Upto speeds have been used in playing marketing games about size.  When really what matters is how long it takes to download an email with a 1MB attachment, or how long will it take for the BBC news website frontpage to download, or how often will a webpage or email download timeout?  It's definitely something that an independent agency, such as JD Power and Associates or Consumer Reports, should be evaluating and comparing across operators as mobile access to the Internet becomes as pervasive as mobile voice services. In an ideal world it would be nice to see a number of measurements around a region, for example NYC metro area, comparing operators' performance with mean and percentage of measurements above some 'adequacy' user experience level.

UC (Unified Communications) brings together the communication and data networks of an enterprise.  The Innovative Communications Alliance (ICA), led by Nortel and Microsoft, is creating the hype surrounding UC.  The purpose of this article is not to discuss UC within the enterprise, rather look at the emerging mobilization of UC and the impact it could have on mobile operators. 

North American Mobile Office revenues are predicted to reach $6-7B in 2011, their current CAGR (Compound Annual Growth Rate) is about 40%.  In Europe, Vodafone's Mobile Office service has done well with roughly 2M customers.  As 'human latency,' that is the time it takes for decisions to be made, becomes increasingly important to the performance of businesses.  Enabling employees to be easily contactable and have access to the enterprise applications and data regardless of location is becoming mission critical.

Looking at the benefits of UC mobilization for the three main actors:

• For customers, mobilizing UC helps in removing some of the communications complexity, e.g. multiple voicemail boxes, multiple messaging platforms, multiple identifies and phone numbers.  Doing a quick tally, I currently have 5 mobile numbers (a US postpaid mobile and prepaid SIMs for the UK, Singapore, UAE and France), an office number, 6 IM/VoIP accounts, and 6 email addresses that I know of.  There are simply too many choices and not enough success with today's disparate communication tools.
• For enterprises, UC helps reduce human latency, enabling employees and partners to get in contact as soon as possible to speed up decision making, improving the company's competitiveness and efficiency.
• For mobile operators: mobilizing UC is strategically important as voice becomes just another application on enterprise IP networks, so they must deliver the UC solution else risk significant voice and VAS (Value Added Service) revenue erosion.

Some of the trends impacting this emerging landscape are:

• Enterprise vendors are using the ubiquity of IP to make communications just another "application," e.g. Avaya, Cisco, IBM, Microsoft, Nortel and Oracle.   Microsoft Communicator Mobile will present a significant threat as the corporate directory, messaging, presence and potentially VoIP could be driven 'over-the-top' of the operators network.
• Ultra-mobile PCs and increasing broadband wireless speeds will make the mobile channel 'more' transparent to IP based communications.  Blurring the difference in experience between a 'phone on the go' and the 'PC in the office.'
• Existing mobile email vendors are moving into the UC and enterprise application space, e.g. RIM.  Enabling the Blackberry junkies to use the device beyond email.
• Hosted enterprise software, e.g. Salesforce.com, has achieved broad market acceptance, creating opportunities for all players to mash-up integrated enterprise solutions.

In this emerging environment what is an operator to do?  Recent enterprise market studies show they still prefer to look to the operator for mobilization solutions.  But given the size of the market and the number of innovative players all trying to leverage their incumbency in the customer's experience, the current preference may not be around for long.

• Partnering with the UC vendors is essential, even through there is a potential long-term risk of from UC vendor's handset clients, UC is going to happen within the enterprise first, and be mobilized second.  There is no option but to work with the incumbent UC vendors. 
• For converged operators with an established enterprise solution business, offering a hosted UC solution with mobilization will be attractive to a number of enterprise segments.  Though an emerging issue will be the integration of this with enterprise applications and data, that is being able to mash-up the hosted solution into the enterprises' data and processes.
• For mobile-only operators the key is to focus on having a best in class mobile solution that can work across a number of UC platforms.  For example, having the UC client and other mobilization application driven from a common user interface.
• Regardless of operator, the over-riding need is to deliver a tight integrated customer experience.  Offering and charging for separate services, e.g. presence, collaboration, messaging, email, directory, security, and enterprise integration will ensure an enterprise goes elsewhere.
  

I attended the FierceMarket's IMS Executive Summit 2007 on the 19th and 20th September.  The agenda included keynote presentations from Stuart Elby, Vice President, Network Architecture, Verizon; Mark Kaish, Vice President of Voice Development and Support, Cox Communications, and Narothum Saxena, Senior Director of Advanced Technology and Systems Planning, US Cellular.

Learning from the event

The CDMA and UMTS networks have clearly diverged on IMS.  Comparing this conference to IMS Asia in March of this year, the difference could not be starker.  The reason is CDMA operators decided voice over EVDO (Evolution Data Optimized) will be VoIP, and IMS provides a standardized service control layer to support VoIP and other IP multimedia communication services.  UMTS does not require VoIP to implement voice services, so the drive does not exist.

Notes from the event

Stuart Elby's Keynote: An open and frank presentation on the fact that VZ's (Verizon) motivation for IMS is long-term operational cost savings in the services layer.  Highlighted the critical issues of:

• Inter-carrier peering to enable VoIP and other IP-based multimedia services to work across operators.  Which is critical to meet the Internet threat as services must work with my contacts on other networks; and
• ANI (Application Network Interface) to enable application developers, enterprises and customers to harness the capabilities in the network such as customer context, authentication, calling and messaging; to do stuff they want.

VZ plans to deploy CSCF (Call Session COntrol Function) and HSS (Home Subscriber Server) components in '08.  Stuart highlighted the critical role of the SDP (Service Delivery Platform) in exposing the capabilities of IMS and other assets in their network.  This was one of the few presentations that had a cogent position on IMS and SDP, while most either ignored the SDP or thought Parlay was perfectly adequate; I'll do an entry on this issue soon.

During questioning the issue of how enterprises use IMS capabilities in an operator's network was raised, for example most of the flows will appear as encrypted VPNs (Virtual Private Network) because the enterprise wants to protect its information.  This is an area that needs attention, because enterprise communications are becoming a critical battle ground, witness Cisco's Smart Business Communication Solutions.

Mark Kaish Keynote:  Mark gave an inspirational presentation.  Critical issues for Mark are bundling is becoming commoditized, and the ability to prime the pump with compelling applications for the many sub-segments in their customer base.  He highlighted the importance of using the Strategic Services Roadmap to drive the case for action in an operator.  They're currently running an IMS technical trial for 85 of their employees, what has impressed them most is the speed at which new services can be created and deployed.  He gave examples of a day planner and a network address book.  During the questioning, the issue of how an operator markets so many applications was raised.  Here Web 2.0 comes into play in using community / viral marketing, as well as good old bundling and segmentation.  Mark also highlighted a problem common to many IMS trials, that SIP stacks are not all the same, they had handset software issues to address during the trial. 

Narothum Saxena Keynote: US Cellular plan to trial IMS in '08.  They view IMS as not an 'if question', rather a matter of 'when.'

Session on Device Roadmap: Paul Callahan (Airvana) highlighted for me the key point of the conference, because EVDO revA mandates VoIP Sprint and Verizon Wireless must adopt IMS much earlier than their UMTS colleagues.

Session on Killer Apps for IMS:  General consensus that there is no Killer App, rather a Killer Environment.  But take the video share service as an example, an operator can implement it for $200k as a silo, or $1M with a lightweight IMS framework.  Simple economics drive operators' decisions to not adopt IMS.  Femtocells were also a 'flavour of the month' driver for IMS, c.f. presence, mobile instant messaging, push to talk and fixed mobile convergence in the past.  I've been requested to do an entry on femtocells, I'll get around to that soon.

Session on What is the ROI for IMS? No ROI numbers mentioned, apart from Ericsson's mentioning they have 40 IMS contracts and 80 IMS trials :) 

Session on IMS Case Studies: Sprint made an insightful comment on using IMS as guidelines, and highlighted the importance of not having architecture limit service deployment decisions.  This echoed a similar point mentioned by Stuart Elby on their drive for A-IMS.

In Conclusion: the drivers for CDMA operators on IMS are different to UMTS operators.  I recommend suppliers with IMS (service control) components or solutions focus upon CDMA operators as the best chance of revenue in the next 18/24 months.  And for UMTS operators outside North America, suppliers not focus on IMS rather upon service innovation and service creation (SDP / SOA (Service Oriented Architecture)).

FMC is a tired term.  Some operators and quite a few enterprises are fed-up of listening to ill-formed, technology-centric FMC pitches.  Taking a simplistic view, when a consumer has a mobile phone, why do they need FMC?  With virtually unlimited 'night and weekend' rates they already have cheap calling at home.  For an enterprise, when their employees have a mobile phone, they now have potentially uncontrolled costs, especially for roaming charges.  So I'm going to look at FMC from an enterprise perspective, because they have a pain point that FMC could potentially solve.  The lack of success of BT Fusion in the consumer market is a data point to back-up this assertion.

Looking at the market there appears to be three broad approaches to FMC:

• The handset centric dual-mode approach where the enterprise replaces the mobile network with short-range radio (WiFi, DECT or Bluetooth) in the office.   They retain the mobile network for off-premises mobility and replace the desktop phone with a dual-mode device. The enterprise retains in their PBX and some voice-over-IP (VoIP) capability.  Here the device vendors, fixed operators and PBX vendors position their solutions.
• The second option is where the enterprise uses signalling and/or VPN (Virtual Private Network) functionality to give it more control over call routing and costs.   It retains its PBX (or hosted PBX), its desktop phones and its mobile devices.  Here the fixed operators, system integrators and also the PBX vendors position their solutions.
• The third, and the mobile operators favourite, is the substitution approach.  Where the enterprise replaces desktop phones with mobile phones and replaces the PBX with a mobile centrex/hosted PBX service.  Here the mobile infrastructure and device vendors position their solutions.

An example of the 'dual mode' approach is BT Fusion Plus.  The Plus is for WiFi rather than Bluetooth.  The BT Business Hub provides the wireless link between BT Fusion handsets and the BT Broadband connection and supports up to 10 Fusion handsets with up to 5 making calls simultaneously.   The Hub is able to connect up to ten PCs, laptops, and peripherals to BT Business Broadband.  They claim savings of about 80% compared to other mobile providers when making calls through the Hub.  For a small business the motivation is simple, communication savings when they're in the office or connected via a BT WiFi hotspot.  FMC is just an element of the solution BT provides, and customers would use it to save money, I'll come back to this point later.

For the 'integration' approach, at the simple end of the scale, the enterprise can simply purchase a bundle of fixed and mobile services from a service provider who is willing and able to provide both, for example, France Telecom's VPN Unifie bundle.  Or Avaya has 'Extension to Cellular,' which is positioned as a feature within Communications Manager, their IP PBX software.  Where software on the mobile phone, Avaya's  one-X Mobile Edition that runs on Nokia Series 60 / 80, Windows Mobile and Blackberry. The client can work out how best to optimize call routing.

And finally the simplest and most successful form of FMC, substitution, otherwise know as Fixed Mobile Conversion.  The enterprise replaces the desktop phones with cellular devices, and replaces the PBX with a mobile Centrex/hosted PBX service. This enables the enterprise to have both fixed and mobile numbers (regulation permitting).  A good example is Vodafone's Wireless Office with roughly 2M customers.

I find it easier to think in pictures, so here is a diagram that gives a fuller view of the landscape I described above.

So given this landscape what are the emerging trends? 

Most enterprises view the 'integration' and 'dual mode' approaches as unproven technologies.  The adoption of multiple new technologies simultaneously is challenging.  Some FMC installations require voice-grade Wi-Fi upgrades; most deployments require all users have a smart phone, and nearly all deployments use VoIP.  Currently there is proprietary and non-interoperable hardware.  High upfront capex costs for devices and hardware as well as a lack of standards and interoperability dilute the cost savings of least-cost call routing.  When IT decision makers think convergence, it isn't about FMC; their convergence perspective is around voice and data (IP telephony).

Bottom line: FMC remains a specialized application in the cold-light of an enterprise's priorities.  A recommendation is to move away from trying to sell just FMC, rather focus upon using FMC as a stepping stone to managed services contracts, e.g. Cisco's SmartBusiness Communications System is a good example of this approach; and focus upon a broader enterprise unified communication and collaboration proposition.  The phasing of this bundled approach must to be driven by customer need, not technology availability.

About 2 years ago I did a couple of projects around the SCIM (Service Capability Interaction Manager).  It's an illusive component within the IMS (IP Multimedia Subsystem) standards, because it's only described in one paragraph of the 3GPP R7 (3rd Generation Project Partnership Release 7) standards issued earlier this year.  The SCIM isn't standardized, it's mentioned.  The SCIM enables service logic across multiple application servers to be combined, e.g. a customer is on a video call and during the call wants to record the voice for 'speak to text' conversion.

As I read the current marketing materials appearing on SCIM you'd think it was a critical mature component, essential to implementing any type of Web2.0 mash-up within an IMS framework.  A Web 2.0 mash-up is combining data and applications from different sources to create a new application.  If you want to find out more, check out some examples at this Web 2.0 list website.

In examining the SCIM market you find three groups of suppliers claiming they do it, i.e. application servers (e.g. Ubiquity), stand-alone SCIMs (e.g. Convergin) or S-CSCFs (Serving - Call Session Control Function) suppliers (e.g. Ericsson).  And then when you scratch the surface a little more you find another three adjacent groups: i.e. legacy services interaction managers, service brokers and SIP (Session Initiation Protocol) service interaction managers.  Here you find suppliers such as AePONA and jNETx.  So is the SCIM a software module, an independent network element, a specialized application server, and is it IMS / IP / SIP / legacy centric?

Many IMS courses and books do not even mention the SCIM.  If an operator has a single vendor application server supplier they may not require a SCIM.  However, it would appear essential if the operator has a switch and SCP (Service Control Point) from different vendors, or if the oeprator has an existing SIP infrastructure they need to migrate to IMS, or if they want to re-use their legacy service logic, or if the operator wants to use 3rd party SIP applications.

In addition the SCIM can be used for conditional service invocation. For example, your schedule says you're in an internal company meeting, so when you're in a meeting call completion is dependent upon who is calling (priorities could be set in either your user profile or address book).  The SCIM enables multiple data sources to be brought into the routing decision, making your communications work better.  The SCIM can also intercept messages to implement different classes of service (business and economy), and enable convergence between Service Management and the SCIM; enabling customers to self-provision new capabilities and services.  The SCIM has a much broader scope that just enabling service logic across different application servers in IMS.

The SCIM would appear to be at the nexus of legacy infrastructure, the SDP (Service Delivery Platform) and IMS.  Should its standardization sit within 3GPP, or within a supplier group such as the SDP Alliance, or should multiple incompatible solutions be produced to let the market decide?

Standards bodies are creating reams of complex standards, IMHO standards engineers should only ever serve a term of one year to avoid standards being created to keep themselves in a job.  Some standard are way ahead of the reality in operators networks, e.g. some operators are struggling to upgrade their networks to 3GPP R4, while R8 is currently being defined.  Earlier in my career I helped set up the FSAN (Full Service Access Network) initiative.  A small group of operators that successfully specified the PON systems being deployed today by operators such as Verizon for their FiOS deployment. 

My recommendation is a small group of operators need to get together and define a common strategic services roadmap to set the requirements, and based on those requirements specify a vastly simplified services layer.  Else operators are currently running a serious risk of missing the last best chance to avoid commoditization into bit-pipes.

The telecom and IT industries tend to suffer more than most from the hype cycle.  What surprises me is the hype tends to be around technologies rather than solutions.  I guess it's easier to get excited about a something that 'could do' rather than 'can do.'  At present the hype machine in gearing up on Web/Voice/Telco 2.0, mash-ups and the threats/opportunities this presents to operators.

I'm seeing lots of arguments about what constitutes a mash-up, or a Voice 2.0 versus Voice 1.5 application.  For those of you who used dial-up message boards on 9.6 kbaud modems in the late '80s and early '90s, will perhaps see Web 2.0 as community and sharing coming back into fashion.

But my objective in this entry is not to worry about definitions.  Web/Voice/Telco 2.0 is here, it sort-of makes sense as a category.  So let's focus on understanding what it is in a little more detail, paraphrasing Sun Tzu, know the lay of the land before entering the battle.  On mapping out the Web 2.0 environment, the positioining I use is it's about "helping us do stuff."   There is a great site that lists most of the 2.0 companies.

The taxonomy I use has five categories: of course Community is at the centre, with Tools, Business, Content (both commercial and UGC(User Generated Content)) and Fun around that centre.  For an operator the opportunities reside on the overlaps between Community and Business, Community and Tools (in particular the communications, identity and geo-location sub-segments), and I've noticed some gaps in the taxonomy of relevance to operators in security and age-rating.  I'll share more on this model and what it means to an operator's in later entries.

Looking out at what operators are doing today, exploration is taking a number of forms.  Three in the UK has enabled access to a number of Web 2.0 applications from its X-Series devices.  A relatively closed system approach, but it does enable both 3 and its customers to experiment in using these applications.  And provides a clear business driver for the operator, simply encourage some customers to buy a data package to access their current Web 2.0 obsession.  BT has taken more of an "open architecture" approach, and with the BT 21C SDK, allowing developers to experiment in creating applications through its ANI (Application Network Interface).  As always with BT, a more techy-driven approach, but an industry leading sandbox nonetheless, which has the potential to define an ANI that most operators could adopt.

Singapore is running an interesting experiment with its IN2015 (Intelligent Nation 2015) initiative.  Quoting from their website:  "The Next Generation National Infocomm Infrastructure (Next Gen NII) is Singapore's new digital super-highway. Next Gen NII comprises complementary wired and wireless networks to ensure Singaporeans enjoy seamless connectivity.  The wired broadband network or Next Generation National Broadband Network (Next Gen NBN) will deliver ultra-high broadband symmetric speeds of 1Gbps and above, to all homes, offices and schools, while the Wireless Broadband Network (WBN) will offer pervasive connectivity around Singapore." 

The plan appears to be creating an 'optical aether' to enable a separation of services from connectivity.  Simply, a service provider doesn't need to worry about the mechanics of reliably streaming a 2 hour HD movie at 16Mbit/s to a customer, because the complete movie is delivered within a few seconds.  In reality the core network is the choke point, as router capacity is not there yet....

On the business structure, the Next Gen NII is an attempt to make broadband access a utility like electricity, gas and water.  Their plan is to have a Consortium (under a PPP, Public Private Partnership) run the network and offer a flat rate connection model for customers, at a similar price to current DSL.  The consortium will NOT act as a service provider over that network.  Of course this is where the business model becomes interesting; to offer a completely new fibre infrastructure throughout Singapore, most probably based upon GPON (Gigabit Passive Optical Network), and make money at a price-point of DSL requires significant innovation.

In Singapore's favour, is an attractive business market, which is still dominated by SingTel, i.e. there's 'margin' available in the business telecom segment.  It has extensive modern utility infrastructure which can in-part be re-used.  And its population density is ideal for just-in-time deployment architectures and technologies. 

A number of business models have been discussed, the chosen one appears to be the Consortium acts as a wholesale provider and retail service providers rent access for their customers (wholesale model).  The analogy I draw is to LLU (Local Loop Unbundling), which has worked very well in Europe, and in particular France where triple-play prices are one third of the US.

However, another option is where the Consortium provides broadband access to customers, and then customers choose their service providers independent of their access provider (retail access model).  Using a utility analogy, as the electricity provider to your TV does not provide the services you watch on that TV, nor should the 'utility' broadband connectivity provider to that TV have a monopoly over those services.

Either way it will be an interesting experiment, and possibly worth testing both business models, where one Singapore region uses the wholesale model, while another uses the retail access model.