1. Market implications for different
deployment models of a central
cognitive enabler
Pieter Ballon IBBT-SMIT (Vrije Universiteit Brussel)
1st IBBT-MIT Joint Workshop on Cognitive Radio Standardization
and Markets, Brussels, Belgium, 11 May 2009
Studies on Media, Information & Telecommunication

2. Research Question
 CPC can have various implementations
 How to evaluate business viability a priori?
 distribution of roles over actors cannot be taken as a given
 assumptions needed for business case cannot be made yet
 Architectural (re)design involved, also at industry level
 Context of platformisation of mobile industry
 In this case: business model analysis is appropriate

3. The Business Model Construct - Definition
 Definition
 Network of Firms
 Specifies control and
value
 In context of
reconfiguration of both
technical and business
architectures
 Functions
 Vertically bridging
 Horizontally bridging

4. The Business Model Construct - Operationalisation
1. Business model archetypes around gatekeeper role(s)
2. Business Model Configuration Matrix: crucial parameters and trade-offs
3. Strategic fit within contextual contingencies and power asymmetries
CONTROL PARAMETERS VALUE PARAMETERS
Value Network Functional Financial Model Value Configuration
Parameters Architecture Parameters Parameters
Parameters
Combination of Assets Modularity Cost (Sharing) Model Positioning
Concentrated Distributed Modular Integrated Concentrated Distributed Complement Substitute
Vertical Integration Distribution of Revenue Model User Involvement
Intelligence
Integrated Disintegrated Centralised Distributed Direct Indirect High Low
Customer Ownership Interoperability Revenue Sharing Intended Value
Model
Direct Intermediated Yes No Yes No Price/ Lock-in
Quality

5. CPC as gatekeeper role
 Solves lack of information in flexible spectrum context
 Adds value by facilitating seamless network selection
and access
 Is bottleneck because it controls and coordinates
information

6. CPC archetypes (1/3)
 Operator CPC: operators deploy their own CPC. They control the
parameters for the information that is offered by the CPC as well
as the usage policies, and own the network over which the CPC
information is transmitted;

7. CPC archetypes (2/3)
 Intermediary CPC: one or more CPCs are operated by a non-
operator entity. A public organisation such as the regulator or new
business actors could take up the role of providing a CPC

8. CPC archetypes (3/3)
 Hybrid CPC: a general ‘meta-CPC’ is operated by either the
regulator or an intermediary and refers to lower-level, individual
CPCs deployed by the operators, within a hierarchical system. The
meta-CPC can be non-exclusive or exclusive

9. Interview Approach
 The interviews were in-depth and semi-structured, and were composed
of a set of qualitative and open-ended questions
 They were conducted by telephone and took generally between 1 hour
and 1 hour and 30 minutes
 Interviewees were sent a paper on the CPC concept and configurations
identified earlier, as well as an indicative questionnaire, prior to the
interview
 The business model configuration matrix was used to structure the
analysis around the most relevant parameters and trade-offs

10. Interview Questions (1/3)
A. Value network
1 W h o should operate the CPC?
2 W hat are drivers and bottlenecks to operate the CPC?
3 Can a CPC function as an independent commercial company? Why (not)?
4 Will the owner of the CPC have a direct relationship with customers?
5 If the CPC is operator- or intermediary-based, can a user change from one CPC to
another?
B. Functional Architecture
6 Should the CPC be integrated into the standard telecom infrastructure (e.g. as a logical
channel within an existing RAT) or should it be separate from the existing
infrastructure?
7 W hat kind of data can be sent from and to the CPC? Possibilities:
7 . 1 F r o m operator’s side:
Available RATs, Capacity, Bandwidth, QoS parameters, Pricing
7 . 2 F r o m customer’s side:
Identification, Location, Desired service class, Device class/brand/OS
etc., Desired QoS/bandwidth/price
8 W here should the decision-making on CPC-enabled service discovery lie?
8 . 1 U s er (active)
8 . 2 D e v i ce (policy)
8 . 3 C PC (brokerage function)
8 . 4 O p erator (operator-originating policies, brokerage function)
Should the bearer for different CPCs be standardized for all operators?

11. Interview Questions (2/3)
C. Financial Model
1 H o w do you estimate the site cost and operational cost of a CPC network? Given the
difference between broadcast/on demand, different bearers, logical channel vs. separate
CPC, spectrum sharing for different CPCs,…
2 H o w do you estimate the cost of CPC alternatives?
3 H o w do you estimate the capacity gain due to CPC-enabled Flexible Spectrum
Management?
4 Can we expect consumers to pay for CPC-enabled services?
5 Can we expect operators to pay for being present onto an intermediary CPC?
6 Is there opportunity for indirect revenues via the CPC, either by government subsidies of
via advertising?
7 If the CPC is operated by an intermediary, what kind of revenue sharing agreements
could be envisaged?
8 W hen do financial transactions take place?
8 . 1 W hen a RAT is listed onto a CPC?
8 . 2 W hen a RAT gets priority listing on a CPC?
8 . 3 W hen a consumer makes use of a CPC to consult the parameters of a RAT?
8 . 4 W hen a consumer subscribes onto a CPC/brokerage function?
8 . 5 W hen a consumer makes a connection to one of the listed RATs?
a. W hen a consumer sends and receives data via one of the listed RATs?

12. Interview Questions (3/3)
D. Value Configuration
1 W hat kind of CPC-enabled, consumer-oriented services may be envisaged?
5.1 Choice of multiple, competing (substituting) or complementary RATs
5 . 2 A lways best connected schemes
5 . 3 A lways cheapest connected schemes
5 . 4 Other
2 W hat kind of CPC-enabled, operator-oriented services may be envisaged?
3 How would these products be positioned vis-à-vis existing services?
4 To what extent will a CPC leave choice for users to select operators and RATs freely,
and to what extent will this be regulated by
4 . 1 U s er defined policies within the terminal
4 . 2 C PC/broker strategies
4 . 3 O p erator lock-in strategies
4 . 4 O p erator lock-in strategies + operator defined policies
5 In the products offered to consumers, what strategy should be chosen?
5 . 1 O p erational excellence
5.2 Product leadership
a. Customer intimacy

13. Interviewed experts
Title Organisation Countr y Date of interview
Mobile Operators
1. Senior Expert, New Telefonica Movile Spain 29 July 2007
Network Technologies
2. Services Sciences France Tele c o m France 30 July 2007
Research Group Co-Leade r
3. Business Manager, Radio Telefonica Movile Spain 31 July 2007
Network Planning
4. Technology Strategy Proximus Belgium 15 Sept 2007
Manage r
5. Head of Regulatory Proximus Belgium 15 Sept 2007
Affairs
6. R&D Project Manager Telecom Italia Italy 24 April 2 0 0 8
Telecom Vendor s
7. Global Marketing Motorola UK 12 July 2007
Manage r
8. Senior Specialist, Nokia Siemens Germa n y 26 Sept 2007
Principal Engineer Networks
9. Product Manager LG France 26 Sept 2007
10. Standardisation Engineer LG France 26 Sept 2007
11. Manager, Research Alcatel Lucent Germa n y 19 Dec 2007
Departme n t
Regulatory and Competition Expe r t s
12. Senior Member Tata Consulting Europe and 29 July 2007
Services India
13. Managing Direc t o r T-R e g s Belgium 19 Oct 2 0 0 7
14. Senior Consultant – IDATE France 30 Oct 2 0 0 7
Head of Mobile Broadband
15. Managing Direc t o r SFC UK 4 April 2 0 0 8
16. Senior Fellow Centre for Belgium 11 April 2 0 0 8
European Policy
Studies

14. Strategic Fit assessment (1/12)
Operators Vendors Regulatory experts
Combination Concentrated Concentrated Concentrated
of Assets
Operators control Operators are wel l Any large‐scale CPC
crucial assets in equipped to set u p implementation
terms of legacy their own CPC at a needs to leverage
networks, fraction of the t ime existing
equipment and sit e and cost that it infrastructure as
ownership. Also, woul d take much as possible ,
mobile operators alternative thereby creating a
control a range of providers. They major advantage fo r
confidential have the le verage existing large mobile
informatio n on to influence policy operators. Also, the
network makers as well as Operator CPC model
deployment, local to acquire contro l would create most
capacity, pricing over any competing value to larg e
and access network independent CPCs operators owning
usage that is hard multip l e RAT
or impossible to networks
obtain without
their consent

15. Strategic Fit assessment (2/12)
Vertical Integrated Partially or Fu lly Partially or Fully
Integration Integrated Integrated
The CPC should be The Operator CPC is In first instance, an
an instrument that feasible, but in some Operator CPC may
safeguards cases the Hybrid CPC be implemented.
operator control model is explici tly Afterwards, more
over spectrum preferred over this. collective models
usage within a F SM Some also consider can be expected. The
context, rather than that the Operato r Hybrid model, with
to further e rode CPC would allow a consortium
this control. The complementary operating the meta‐
Operator CPC i s operators or that an CPC, is advocated b y
preferred. Some operator consortium most competition
‘closed platform’ would establish a experts
variants or a ‘th in’ joint CPC
Hybrid CPC, whil e
not the preferred
options, are seen a s
potentially feasible

16. Strategic Fit assessment (3/12)
Customer Direct Direct Direct
Ownership
Users will not be Most end‐users do Operators should be
able to sca n not value a free to i mplement
different operator brokerage function their own busines s
CPCs, nor t o switch in order to actively models. No
between operators switch between agreement as to
in a dynamic operators. Some whether user
manner experts envisage terminals would be
that end‐user able to ‘see’ different
terminals will be CPCs when
able to ‘see’ different connecting to a
CPCs when meta‐CPC
connecting to a
meta‐CPC

17. Strategic Fit assessment (4/12)
Modularity M o d u l a r , S e m i ­ Semi­Modular Semi­Modular
Modular o r
Integrated
Difference of The optimal trade‐ What is advocated is
opinion or off may be either an an in‐band CPC tha t
indecision as to in‐band solution makes use of
whether the CP C that incorporates operators’ existing
should be an ‘out ‐ some form of network
band’ channel, modularity, e.g. by infrastructure,
independent of designing the CP C possibly (and in
legacy systems and as an application most cases,
RATs, or an ‘in ‐ server that could be preferably)
band’ logical separated from the combined with a
channel within network meta‐CPC that
existing systems management makes use of
and RATs domain, or as a existing or
Hybrid CPC, with a dedicated
general out‐band infrastructure of the
channel being meta‐CPC
combined with an consortium
operator‐specific members
channel that
contains most of the
CPC functionality

18. Strategic Fit assessment (5/12)
Distribution Mostly Centralised Mostly Centralised or Partly Distributed
of Partly Distributed
Intelligence
The Operator i s The responsibility Most intelligence is
responsible for for optimising QoS located at th e
optimising network and cost should lie operator’s side. The
behaviour. Some with the operators. data on the meta ‐
consider that there Still, the end ‐user CPC side should be
might be generic may be triggered to restricted. There
policies set i n make certain choices may be policy ‐
advance by th e and may be base d network
users for selecting presented with price selection, set by
particular RATs information in t he end‐users
within the form of
operator’s domain advertisements

19. Strategic Fit assessment (6/12)
Interoperability Standardised Some elements Some elements
standardised standardised
The CPC shou ld In case of an out ‐ Non‐standardisation
be standardised band (meta‐)CPC, of the CP C
either as stand ‐ the bearer technolo g y may
alone solution or network should cause the reach and
inside a cellular be standardised, comprehensiveness
technolo g y . A at least on a of the CPC to b e
Hybrid CPC cou ld European level. restrained. Yet
accelerate In case of an in‐ standardisation risks
introduction, by band CPC, only to diminish any
only the detection advantages of inter‐
standardising the procedure should technology
meta‐CPC be standardised competition . A
Hybrid CPC model
might strike the bes t
b a l a n c e between
both considerations

20. Strategic Fit assessment (7/12)
Operators Vendors Reg. experts
Cost Mostly Concentrated Mostly Concentrated Partly Distributed,
(sharing) Partly Concentrated
model
Cost of Operator CPC Cost of a CPC i s Costs are partl y
is relatively light f or relatively low if distributed as far as
established existing a consortium model
operators, high f or infrastructure is used is advocated, and as
others and/or if a n partly concentrated
application server regarding the
logic is employed individual operator
investments in an
Operator CPC

21. Strategic Fit assessment (8/12)
Revenue Indirect Indirect Indirect
model
The CPC is pr imarily The CPC can function Indirect revenues are
a way to incre ase as an advertising and generated through
efficiency and thus to marketing channel. cost reduction, better
cut costs and Also, it may allow quality of service, th e
increase profitability. operators to avoid a facilitation of
In addition, new part of their network additional services,
connectivity bundles costs, and enable and an ex pansion of
could be proposed, them to more opportunities to use
within a flat fee optimally distribute existing services.
revenue model logic their end‐users over Versioning of flat fe e
the various rad io packages is also
access networks envisaged

22. Strategic Fit assessment (9/12)
Revenue Yes, no specific role Yes, with limited role Yes, with role for CPC
sharing for CPC for CPC
model
National roaming Revenue sharing may Revenue sharing
agreements would take place as a result may take place as a
mirro r current of roaming and/or as result of inter ‐
international a result of usin g operator roaming
roaming agreements other operators’ CPC agreements and of
infrastructure sharing CPC capacity

23. Strategic Fit assessment (10/12)
Operators Vendors Reg. experts
Positionin Complement Complement Complement
g
The CPC is mainly an The CPC may be mainly The CPC, rather than t o
enabler to mana ge and invisible to end‐users, in lead to stron gly
control heterogeneity, an d terms of bot h disruptive changes, will
optimise network efficiency.
functionality and enable a series of
Market positioning would
not change at all or only
positioning of operators. opportuniti e s for
slightly by offering mor e Alternatively, an ‘always differentiating the
user choice and favouring best connected market positioning of
large operators guarantee’ could o p e r a t o r s , t h r ough
constitute an additional versioning, facilitating
selling point for current new services and so on
operators. The CP C
could also be visible to
users in the form of an
informatio n or
marketing channel

24. Strategic Fit assessment (11/12)
User Limited or non­existent Limited or non­existent Limited
involvem
ent The CPC functionalities Most or all of the Some doubts as to
should be as ‘ invisible’ as responsibility for whether end‐users will
possible, a.o. because of the optimising n e t w ork have the possibility to
shift towards flat rates for
parameters should lie ‘see’ different CPC. Even
connectivity, and th e
negative influence on use r
with th e operators. in the case that ‘active ’
value of custom ers having Possibly, a restricte d switching between RATs
to make frequent network number of relevant and/or providers is
connectivity choices. There instances could be allowed, this should take
could possibly be some us er filtered out for which place as a r esult of
pre-sets the user is triggered to policies of which some
make an active choice are set in advance by the
operator and others by
the end‐users

25. Strategic Fit assessment (12/12)
Intended Mixed Mixed Mixed
value
Spectrum efficiency Through the Customer intimacy may
enhancing tools such a s optimisation of radio dominate in the short
the CPC should a llow access network term, while optimisation
optimisation of the selection, end‐users of the price/quality ratio
price/quality ratio, but could be offered either may dominate in the
these advantages may the guarantee of bein g longer term. Various
be mostly absorbed by ‘best connected within opportunities for inter‐
operators. The limited certain cost constraints’, operator differentiation
user choice, th e or of enjoying very are foreseen, because of
provision of flat fee cheap free connectivity. divergent CPC
connectivity packages In addition, the CPC is implementations and
and the single operato r likely to be employed to opportunities to offer a
control over the keep end‐users within a range of additiona l
information provided by single operator’s services
the CPC promote a domain
customer intimacy o r
‘lock‐in’ strategy

26. Validated CPC business models (1/3)
 Operator CPC: all operators have their own CPC. Operators control most of the value
network and technical and customer-data on the CPC
 Users switch (passively or actively) between their home operator’s available networks but
continuously remain within the same business domain. Any technical platform activity in
case of spectrum trading is transparent
 Pre-established revenue sharing models set the rules for clearing between the network
operators (national roaming)

27. Validated CPC business models (2/3)
 Extended Operator CPC: CPC functionality is opened up to other connectivity providers
that do not possess a CPC infrastructure. This is a platform in a business sense because
users can access different business domains
 However, Operator CPC platform not available to every stakeholder. Mainly complementary
operators: in terms of network reach (i.e. smaller operators with niche technology or with
network deployment in a limited area), or in terms of customer reach (i.e. MVNOs with
attractive branding)

28. Validated CPC business models (3/3)
 Hybrid Consortium CPC: Information provision for large ones and small
operators, while operators retain control over own pilot channel and over
own connectivity offering. From the 3 strategically feasible models, this one
alleviates most the regulatory experts’ concerns
 Only one CPC channel needs to be harmonised and needs to be known a
priori by the device. Non-exclusivity of meta-CPC only in case of strong
regulatory intervention

29. Discarded CPC business models
 Intermediary commercial CPC: introduces single point of failure, does not have
necessary control over data, no strategic fit between stakeholders, control of
customers not aligned with control of gatekeeping roles, continuous customer
choice in terms of mobile access decreases instead of increases customer value
 Intermediary regulatory CPC: introduces single point of failure, no strategic fit
with main stakeholders, continuous customer choice in terms of mobile access
decreases instead of increases customer value
 Hybrid regulatory CPC: regulatory body is most likely to outsource meta-CPC
functionality (cfr. Number portability databases etc.)
 Hybrid commercial CPC: no commercial value proposition

30. Conclusions (1/2)
 The need for a CPC-type functionality as well as the likelihood of any major business model
reconfiguration depend on the (uncertain) persistence and intensification of both heterogeneity in
networks and flexible spectrum management
 Even under such circumstances, no strongly disruptive outcome vis-à-vis the current business
model configuration
 Interpretation of the introduction of AWTs and even of FSM as not necessarily
negative for current mobile operators’ businesses,
 Estimation that the assets to compete in the mobile access provision domain, and
more specifically to offer the CPC’s gatekeeping functionalities, are firmly
controlled by established mobile operators
 Expectation that moving (most of) the CPC functionalities away from the operators
would diminish, rather than increase, customer value, because it would potentially
lead to information overload and to unpredictable tariffs for connectivity

31. Conclusions (2/2)
 Towards platform rather than integrated model?
 Platformisation depens on increasing trend to RAT heterogeneity and flexibility of spectrum
 Validated business models include two platform models as well as an integrated, non-platform model. The
first platform model concerns an Extended Operator CPC model in which the operator uses its CPC as a
platform for (mostly complementary) smaller operators and MVNOs. The second one refers to a Hybrid
Consortium CPC model in which a meta-CPC is set up by a consortium of operators, and functions as a
platform providing generic information on the location of individual operators’ CPCs.
 Whether platformisation will in fact take place in such a context, depends a.o. on the cost structure as well
as the cost saving potential of particular CPC implementations, and on the regulatory insistence on a joint
solution
 any further reconfiguration and platformisation beyond these validated models depends on regulatory
intervention that would go significantly further than the current consensus
 Strategic Fit around the value configuration related to CPC and cognitive radio
 Customer intimacy and Operational excellence, Not diversity of offerings is stressed. A CPC will firstly
ensure efficiency and seamlessness; and secondly (in case of the platform models) will safeguard the
existence of niche operators

32. Contact
 pieter.ballon@vub.ac.be