Académique Documents
Professionnel Documents
Culture Documents
School of Science
Degree Programme of Computer Science and Engineering
Rushil Dave
Masters Thesis
Espoo, August 8, 2011
Supervisor:
Instructor:
Aalto University
School of Science
Degree Programme of Computer Science and Engineering
ABSTRACT OF
MASTERS THESIS
Author:
Rushil Dave
Title:
Mobile Virtual Network Operator
Systems on Cloud: An Architectural and Cost-benefit Study
Date:
August 8, 2011
Pages: 9+ 98
Professorship: Data Communication Software
Code: T-110
Supervisor:
Professor Antti Yla-Jaaski
Instructor:
Yrjo Raivio Lic.Sc. (Tech.)
Cloud computing technology is growing rapidly in the market providing elasticity, virtualization and most importantly cost savings to the computing industry.
Today, telecom operators especially MVNOs face challenges such as reduction of
ARPU, need of differentiations and competition from all fronts of software industry. Cloud provides major opportunities to telecom operators to reduce their
operational costs, drive new differentiations and be active in cloud ecosystem.
Telecom operators can take advantages of cloud by selling the cloud offerings to
their customers by using already established infrastructure as well as by using
cloud to deploy their own software systems to reduce operational costs.
In this research, the techno-economic analysis has been carried out to analyze
implementation of telecom systems on cloud in Business Support Systems domain.
The research focuses on application mapping strategy required and financial costbenefits gained for MVNOs if their systems are implemented on cloud. The
thesis research involves Delphi method based study in which expert interviews
are conducted in two rounds to get expert views on the subject. The responses
and related analyses are drawn while introducing a proof-of-concept hybrid cloud
architecture to implement MVNO systems. An innovative solution of billing-asa-service on cloud is discussed in the thesis as well. The implications and future
prospects are presented evaluating the thesis results. In the end, conclusion is
derived to summarize the thesis research.
In summary, most of the MVNO systems are viable candidates for cloud implementation and significant benefits can be achieved. Cloud will help MVNOs to
make transition from CAPEX to OPEX based model while reducing OPEX and
gaining other advantages like economy of scale, SaaS based solutions, lower timeto-market, enhanced performance etc. There are certain implications related to
data security, availability and government regulations which can impose difficulties on MVNO systems while moving them to the cloud. However, the future of
cloud in telecom industry seems bright with emergence of everything-as-a-service
model.
Keywords:
MVNO, BSS, cloud, cost-benefits, hybrid, SaaS, SLA, OPEX
Language:
English
ii
Acknowledgements
This work was done as a part of the Cloud Software Program of Tivit
(Strategic Centre for Science, Technology and Innovation in the Field of
ICT, www.tivit.fi).
I wish to express my gratitude to the people that have supported me
during this thesis research.
Firstly, I would like to thank Professor Antti Yla-Jaaski, who acted as a
supervisor for the thesis and helped me to initiate the thesis research.
Secondly, I would like to extend my sincere gratitude to Mr. Yrjo Raivio,
who was the instructor during thesis research. His guidance, support and
comments helped me a lot throughout the research process to carry out the
thesis research.
I also thank all the experts and researchers for their anonymous participation in the interview process providing expert knowledge and comments.
I would also like to thank Amit Soni and Koushik Annapureddy, who
helped me to setup the CRM proof-of-concept on hybrid cloud and review
the thesis.
Last but not the least, I thank my family and friends for their support.
iii
iv
MSISDN
MSS
MVNE
MVNO
NAS
NIST
OPEX
OSS
PaaS
PHP
RAM
ROI
SaaS
SIM
SLA
SMS
SP
STOF
TCO
VAT
VLAN
VOIP
xDR
Contents
Abstract
ii
Acknowledgements
iii
iv
List of Tables
viii
List of Figures
1 Introduction
1.1 Background of Research .
1.2 Research Questions . . . .
1.3 Research Objectives . . . .
1.4 Research Methodology . .
1.5 Scope of Research . . . . .
1.6 Structure of Thesis Report
ix
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2 Overview
2.1 MVNO Architecture and Systems . . . . . . . . .
2.2 Business Support Systems . . . . . . . . . . . . .
2.3 Cloud Environment and Architecture . . . . . . .
2.4 Existing Scenario with Telecom Systems on Cloud
2.5 Application Mapping and Finance . . . . . . . . .
2.5.1 Application Mapping . . . . . . . . . . . .
2.5.2 Finance . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1
2
3
4
5
6
.
.
.
.
.
.
.
9
9
11
14
19
22
22
24
3 Action Research
27
3.1 Action Research Methods . . . . . . . . . . . . . . . . . . . . 27
3.1.1 Action Research Setup . . . . . . . . . . . . . . . . . . 28
3.1.2 Qualitative Data Analysis . . . . . . . . . . . . . . . . 28
vi
3.2
3.3
3.1.3 Interviews . . . . . . . . . . . . . .
3.1.4 Delphi Method . . . . . . . . . . .
Expert Interviews Setup . . . . . . . . . .
Interview Responses . . . . . . . . . . . .
3.3.1 MVNO Systems . . . . . . . . . . .
3.3.2 MVNO on Cloud . . . . . . . . . .
3.3.3 Cost-benefits . . . . . . . . . . . .
3.3.4 Implications and Future Prospects
3.3.5 Round 2 Results . . . . . . . . . .
4 Techno-economic Analysis
4.1 Mapping MVNO Systems to Cloud . . .
4.1.1 Interview Analysis . . . . . . . .
4.1.2 Application Mapping . . . . . . .
4.1.3 Strategy . . . . . . . . . . . . . .
4.1.4 Hybrid Cloud Setup Experiment .
4.1.5 Analysis of Round 2 . . . . . . .
4.2 Cost-Benefits . . . . . . . . . . . . . . .
4.2.1 Cost Calculations . . . . . . . . .
4.2.2 Opportunity Gains . . . . . . . .
4.3 Parameters Evaluation . . . . . . . . . .
4.3.1 Importance in MVNO Context .
4.3.2 Influence of Cloud in MVNO . . .
5 Discussion
5.1 Billing as a Service . . .
5.2 Business Model Analysis
5.3 Implications . . . . . . .
5.4 Future Prospects . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
29
30
31
34
34
41
48
51
53
.
.
.
.
.
.
.
.
.
.
.
.
55
55
55
57
60
61
65
67
68
70
71
71
73
.
.
.
.
76
76
79
80
83
6 Conclusion
86
Bibliography
87
A Appendix
95
A.1 Questionnaire for Interview Round # 1 . . . . . . . . . . . . . 95
vii
List of Tables
3.1
3.2
3.3
3.4
4.1
4.2
4.3
4.4
60
69
70
70
viii
List of Figures
1.1
1.2
2.1
2.2
2.3
2.4
2.5
2.6
2.7
.
.
.
.
.
.
10
11
12
13
16
17
.
.
.
.
21
22
24
25
MVNO Positioning . . . . . . . . . . . . . . . . . . . . . . . .
MVNO Types . . . . . . . . . . . . . . . . . . . . . . . . . .
Telecom Operator Systems . . . . . . . . . . . . . . . . . . . .
MVNO Business Support System Processes . . . . . . . . . . .
Cloud Computing Structure . . . . . . . . . . . . . . . . . . .
Advantages of Cloud Computing Services . . . . . . . . . . . .
Cloud Implementation of Telco Systems by Carrier Grade Requirements, adapted from [45] . . . . . . . . . . . . . . . . . . . . .
2.8 Telecom Cloud Offerings . . . . . . . . . . . . . . . . . . . . .
2.9 Possible MVNO Applications Mapping on Cloud . . . . . . . .
2.10 Cost Analysis Parameters for MVNO Cloud . . . . . . . . . . .
5
7
3.1
3.2
3.3
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10
5.1
5.2
ix
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
56
57
58
61
62
64
65
68
72
74
Chapter 1
Introduction
1.1
Background of Research
Telecom operators are moving towards the flat network architectures and
sharing of network resources is now possible in this domain by introduction
of Long Term Evolution (LTE) and 4G standards [46]. The flat network
architecture is lowering the barrier to network sharing initiatives also enabling the sharing possibilities of Telecom IT systems. The network sharing
was started taking place in the past when MVNOs [30] were introduced.
MVNOs share network resources with MNOs and they implement their own
Customer care, inventory management and billing systems as part of their
IT systems. Most of the MVNOs are aggressive on the pricing of telecom
products and services offered and price differentiation is vital in MVNO business operations. MVNOs pay large amount of their spending on buying the
network capacity from MNOs as they introduce products and services while
competing against the MNO offerings. So its important for MVNOs to lower
the operational expenditures by lowering the cost of running their IT systems
in place to stay in the highly competitive market.
Now the cloud computing technology [6] has given possibility to host systems which require high processing power and large data storage capacity
in an efficient manner. In addition to that, cloud computing environments
provide possibility of hosting Infrastructure, Platform and Software as service by dynamically setting up the use of different types of cloud. Private
cloud gives possibility to share in-house IT resources efficiently while sharing
with the processes, batch jobs and services in a highly secured and reliable
fashion. Public cloud gives possibility to use public computing resources dynamically and on pay-as-you-go basis which gives very cost-beneficial setup
for processing the complex data. Hybrid cloud combines private and public
CHAPTER 1. INTRODUCTION
cloud computing space for managing sensitive data storage and processing
in-house and consuming the public computing resources whenever there is a
need [64].
In highly competitive market of telecom offerings, every operator is trying
to reduce the operational expenditure in one or the other way and cloud environment gives such possibility to minimize the expenditure and maximize
the revenues. Telecom networks have various functions and network elements
that might be implemented in clouds, called a Telecom Cloud. Operators
can utilize all kinds of cloud computing services, including Infrastructure
(IaaS), Platform (PaaS) and Software as a Service (SaaS). On IaaS layer
communication, computation and storage services can be provided by the
clouds. PaaS layer has close links to Business and Operations Support Systems (BSS/OSS) [40]. Finally, SaaS matches the service delivery platforms.
The Telecom Cloud can be supported with private, public and hybrid clouds
located outside of the operator domain, and separated by open APIs [13].
In cloud environment, however, the mapping of application is also important as to meet with critical SLA and throughput requirements [45] in telecom
sector. Also nature of applications in Telecom IT systems vary significantly
and mapping of applications allows choosing cloud deployable applications
according to their attributes. Mission critical telecom systems may have
impact of the complex telecom operations and using cloud for such systems
may not be a good choice for telecom operators. Cloud provides cost-benefits
while sharing the ideal IT resources with other in-house or public applications, security and availability is a concern for telecom operators as some
of the telecom operators think cloud is not ready for some of the telecom
systems. Its also important for telecom operators to know how and what
cost-benefits can be driven by implementing their telecom systems on cloud
as there factors involved such as application migration, network usage, integration and enhancement possibilities which may impact the operational
expenditure for telecom operators. So the architectural changes and costbenefits are very important factors here to consider while moving telecom
systems to the cloud.
1.2
Research Questions
In the highly competitive telecom services market, every operator seeks possibilities to reduce the costs and to maximize the revenues [58]. The cloud technologies provide several ways like IaaS, PaaS and SaaS to improve CAPEX
and OPEX values by migrating or implementing telecom systems on cloud.
Because MVNOs share various network resources with MNOs, MVNOs can
CHAPTER 1. INTRODUCTION
become natural appliers of clouds. However, the mapping of MVNO architecture to the cloud has to be carefully planned because of the business
sensitivity of MVNO systems. The research questions for this thesis are derived in the direction of MVNO applications mapping possibilities and related
cost-benefits on cloud computing architecture. The main research question
is followed by the sub-questions which are narrowed-down to illustrate important areas of the thesis research.
Main Question:
How and which Business Support Systems (BSS) can be deployed for
MVNO on Cloud computing environment as to offer high performance, crosslocation architecture and cost-effective system to share resources among other
operators?
Sub Questions:
1. Which BSS can be deployed on cloud? (BSS Application Mapping on
Cloud)
2. How will the current architecture be affected in this case?
3. If high-performance will be gained with cloud implementation? How?
(SLA, Throughput in focus)
4. How much cost-benefits can be driven? How much cloud resources
should MVNO buy?
5. What are the implications of BSS cloud deployment? (Like security,
government laws etc.)
6. How much carbon footprints can be saved by using Cloud?
7. How big is the market? What changes will business models have in this
case?
1.3
Research Objectives
The objective of the research defines the criteria and framework for answering
the research questions. It also defines the evaluation criteria for research and
a structure for research while deriving step by step procedure to answer the
research question.
The main objective of the thesis research is to design the mapping of
BSS applications on Cloud for MVNOs and do financial analysis to find out
the cost-benefits related to the implementation. The following objectives
CHAPTER 1. INTRODUCTION
establish the research direction for the thesis while answering the research
questions stated in section 1.2:
1. Define criteria and methodology for the thesis research to map MVNO
systems to the cloud and finding related cost-benefits.
2. Gather qualitative data from industrial and research partners related
to the thesis area of research and analyze it further to derive the results.
3. Conduct architectural and design changes study for MVNO system
and map MVNO systems to the cloud taking qualitative data into
consideration.
4. Calculate cost-benefits while finding impact on MVNO capital and operational expenditures when moving MVNO systems to the cloud.
5. Evaluate parameters like performance, revenue, market, carbon footprints, cloud resources and business model changes for the MVNO systems on cloud implementation.
6. Discuss the solution for providing Billing as a Service for MVNOs using
cloud computing environment using the analysis conducted during the
thesis research.
7. Describe the technical and other implications for the MVNO system
implementation on cloud. Also discuss the future prospects for the
thesis research based upon the input gathered during qualitative data
analysis.
1.4
Research Methodology
Action research will employed as the main research method involving researchers and practitioners together. Expert interviews will be conducted
based upon their area of expertise with pre-formatted and validated questionnaire. The Delphi method [49] will be exercised for the interview formations. The business critical information and knowledge imparted from
interviews will remain confidential and anonymous in the thesis report and
research publications.
Following methods shall be employed during the thesis research for which
the a brief agenda is illustrated in Fig. 1.1:
Literature Review
CHAPTER 1. INTRODUCTION
Expert Interviews
Design/Architectural Study
Statistical/Financial Analysis
Business Model Analysis
1.5
Scope of Research
Scope of the research is limited to the MVNO systems and especially the
business support systems in place for the MVNOs. This is due to the fact
CHAPTER 1. INTRODUCTION
that most of the MNOs have implemented complex in-house telecom IT systems which hold large amount of customer sensitive data. Telecom operators
believe that the cloud is not yet ready to hold such complex, mission critical
and data sensitive systems [38].
MVNOs, on other hand, have systems which are light weight in nature
and hold small amount data and configurations. Some of the MVNOs have
also outsourced their IT systems to the Business process outsourcing companies. So its easy for the MVNO systems to move to the cloud compared to
the IT systems of large MNOs. Also the Operation Support Systems include
network elements and switches which are required to fulfill carrier grade SLA
and in these cases, telecom operators believe that cloud is not yet ready to
hold such systems. So to make the viable case of telecom systems implementation on cloud, MVNOs and their Business Support Systems have been
considered for this thesis research. MVNO in this thesis research includes
MVNE, Service Providers (SP), and resellers in telecom value chain.
1.6
The structure of the report will provide information on how the report will
proceed and what types of different research methods are used to support the
objectives of the thesis. Fig. 1.2 depicts the structure and blocks of thesis
report which are explained as following.
The first part of report starts with Introduction to the report. Introduction also includes research problem, research question and sub-questions,
research objectives, research methods and also the research framework or
design. It also comprises of background and scope of the research being
conducted as part of the report.
The second part mentions some theoretical aspects needed to build the
background of the report. This part consists of several topics supported by
literature reviews and expert interviews. First topic on MVNO systems explains the MVNO positioning in telecom operator environment and different
MVNO types along with MVNO operation and business in brief. Second
topic focuses on the business support system along with its positioning in
telecom IT systems environment and processes involved in MVNO context.
Third topic describes the cloud computing architecture and related benefits
of cloud implementation. Fourth topic discusses the existing scenario of cloud
in telecom system environments. Fifth topic illustrates the theoretical part
for the application and finance related to the MVNO systems implementation
on the cloud.
The third part of the report carries out qualitative data analysis in which
CHAPTER 1. INTRODUCTION
action research methods and the results are described in detail. First subsection mentions different action research methods used in the thesis in detail.
Second sub-section describes the expert interview setup and the Delphi process followed while conducting interviews. Third sub-section demonstrates
the results gathered for the semi-structured interviews conducted as part of
the thesis research.
Fourth part analyzes the gathered results with techno-economic methods in a structured fashion. First topic maps the MVNO systems on cloud
taking views from the action research conducted. Second topic calculates
related cost-benefits taking the cost-structure and pricing information into
consideration. Third topic evaluates different parameters mentioned in the
CHAPTER 1. INTRODUCTION
research questions.
Fifth part discusses different aspects derived from the thesis action research which are related to the telecom systems implementation on cloud.
First section discusses and analyzes an innovative solution of Billing-as-aservice for telecom operators using cloud architecture. Second section analyzes market and business model changes for telecom system on cloud. Third
topic demonstrates technical and other implications while moving telecom
system on cloud. Fourth section discusses future prospects in the direction
of thesis research.
Sixth part concludes the thesis research while briefly reviewing all parts
of the thesis report.
Chapter 2
Overview
2.1
CHAPTER 2. OVERVIEW
10
chain based upon the contracts or agreements [8]. MVNO and MVNE buy
network capacity from MNO while paying the network usage cost on payas-you-go manner. MVNO further provides services to the end-users owning
billing and marketing systems. MVNE sells network capacity and services
[48] to service providers and in some cases to MVNOs. Service provider sells
services to the end-users and in some cases it focuses solely on marketing,
branding and reselling the services i.e. Brand Operator. Fig. 2.1 shows the
classification and positioning of network operator types in mobile network
service providers value chain.
MVNO generally owns services and content delivery, billing, customer
management and marketing systems. In some cases, MVNO also implements
switching and networking elements on its own to have more service fulfillment
possibilities [9]. MVNO has full control in different ways over their inventory,
resource management, SIM card delivery, marketing and branding systems.
Depending on the market share and financial situations, MVNO buy or
implement telecom software elements for itself which differentiates it from
other MVNOs. Below are the types of MVNO [30] based upon the systems
it implements or owns also presented in Fig. 2.2:
1. True MVNO implements HLR, Switching and Intelligent Network
platform on their own in addition to Services, Billing and Marketing.
Such MVNO has control over service and tariff design, service implementation and differentiation and service marketing as well as branding.
CHAPTER 2. OVERVIEW
11
2.2
Fig. 2.3 shows the traditional architecture of telecom operator system [40]
which consists of three layers:
Business Support System (BSS) Layer
Operation Support System (OSS) Layer
CHAPTER 2. OVERVIEW
12
Network Layer
Along with the system layers, operator systems also include integration
enablers to provide the interconnection between layers [21] which provides
application connectivity and solution flexibility. Each of three layers has
specific functions for the operator systems which are explained below:
1. BSS layer: BSS layer focuses towards the customer and financial
transactions segments of telecom operator systems. It also manages
partner and marketing functions of operator systems. The front-end
operations for self-service portal for the end-users as well as for customer service representatives (CSR) are included as part of BSS. BSS
CHAPTER 2. OVERVIEW
13
CHAPTER 2. OVERVIEW
14
or CRM comes into the picture. Subscriber herself or Customer Service Representative (CSR) enters subscriber details along with the price
plan, offers and discount plan are added to the system via CRM. Order
management system then processes the order made in CRM or portal
getting offer related information from product catalog. Order management system then gets information of resources such as MSISDN, SIM,
IMSI, Data number etc from resource management system or resource
inventory. Customer management system then enters customer-specific
details like one time and recurring charges of the offers to the billing
system database. After successful order completion BSS informs OSS
and network systems to activate the subscribers SIM and MSISDN.
2. Processing bills: MVNO gets the Call Details Records (CDR) from
the network operators from whom MVNO has bought the network capacity or radio spectrum. Call detail records hold the information of
the calls (voice and data) being made including the destination, call
start and end time, duration, call type etc. Mediation system gets the
CDRs as an input from different network operator systems, analyses
the data and prepares data in a generalized format for the usage rating
purpose. Rating system rates the calls based upon the prices fetched
from product catalog. Billing system calculates individual charges both
one time and recurring for the subscribers and enter the detail into the
database. Bill formatter gets the data from all BSS systems to prepare
the printable bills and pass it to the distribution channel further.
3. Bill payments: Subscriber pays the bill via MVNO shop or web-shop
or via online payment gateway. BSS has integration with the payment
gateways to receive payments from the subscribers. Account receivable
system takes care of integration with CRM as well as payment gateways
and makes database entry for the payments received from subscribers.
Collection system collects the data from Account receivable and bars
subscriber services if payment hasnt been made until the payment due
date. Collection also informs OSS and network systems to block the
network services for subscriber. After payment received for subscriber,
collection unbars subscriber services again.
2.3
CHAPTER 2. OVERVIEW
15
network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal
management effort or service provider interaction.
Cloud computing [5] is a utility service for computing resources just like
water or electricity where users utilize services sharing with others and paying
the cost on pay-as-you-go or subscription basis. Cloud is a mixture of datacenter hardware and software. Cloud provides benefits to use several server
instances simultaneously for complex processing tasks. For example, large
batch-processing task can be executed using 1000 cloud instances in one hour
which is equivalent one server for 1000 hours.
Cloud computing provides three aspects in form of hardware virtualization point of view:
1. Illusion of infinite processing power for cloud consumers.
2. Elimination of up-front investment in buying server capacity.
3. Ability to utilize computing resources on pay-per-use model just like
the utility services.
As defined by US National Institute of Standards and Technology (NIST)
[6], there are four different cloud computing deployment models (See Fig.2.5)
giving possibilities to use cloud deployment according to the application or
system infrastructure needs:
Private cloud: The cloud infrastructure is operated solely for an
organization. It may be managed by the organization or a third
party and may exist on premise or off premise.
Community cloud: The cloud infrastructure is shared by several
organizations and supports a specific community that has shared
concerns (e.g., mission, security requirements, policy, and compliance considerations). It may be managed by the organizations
or a third party and may exist on premise or off premise.
Public cloud: The cloud infrastructure is made available to the
general public or a large industry group and is owned by an organization selling cloud services.
Hybrid cloud: The cloud infrastructure is a composition of two or
more clouds (private, community, or public) that remain unique
entities but are bound together by standardized or proprietary
technology that enables data and application portability (e.g.,
cloud bursting for load-balancing between clouds).
CHAPTER 2. OVERVIEW
16
CHAPTER 2. OVERVIEW
17
environment configurations.
Software as a Service (SaaS): The capability provided to the consumer is to use the providers applications running on a cloud infrastructure. The applications are accessible from various client
devices through a thin client interface such as a Web browser (e.g.,
Web-based email). The consumer does not manage or control the
underlying cloud infrastructure including network, servers, operating systems, storage, or even individual application capabilities,
with the possible exception of limited user specific application
configuration settings.
There are certain attributes for the cloud [14] which define the architecture and related services on the cloud. These attrbitues are categorized
as following while explaining how they provide various functionalities and
benefits to the IT industry in general:
Cloud Infrastructure provides virtualization of hardware resources. Cloud
infrastructure is simply the virtual servers in cloud. Its beneficial for
large-processing tasks where virtualization is used to reduce time and
CHAPTER 2. OVERVIEW
18
cost significantly. With ability of cloud to provide server instances almost infinitely, it eliminates need to buy server resources and provide
economy of scale by sharing cloud resources.
Cloud Storage is an effective and efficient medium of storing enormous amount of data with pay-you-you-go service model. In this
case, cloud also provides services like data synchronization, Network
Attached Storage (NAS), database-like functionality and also unstructured data services. Data access to large amount of data is possible
anytime anywhere by using cloud storage infrastructure and services.
There are certain limitations due to issues with availability, transaction
support and data transfer bottlenecks in cloud.
Cloud Platform is used to build, test, deploy, and manage applications
on cloud with storage and processing power of cloud itself. Cloud
platforms are low-cost, reliable, accessible anywhere anytime and highly
scalable with provision of web-based services environment. Portability
is an issue because as soon as applications or services start running
on one cloud platform. It becomes complicated task to move systems
deployed on one cloud platform to another cloud platform or back to
the original platform.
Cloud Applications provide Software-as-a-service model which further
eliminates the need to install & configure applications on cloud. Along
with this, it also provides cloud benefits related to virtualization &
storage. Another model is S + S (Software + Service) on cloud in
which cloud hosts use rich client applications interface into an extremely
hosted environment with ability to work in offline mode and synchronize
data when online.
Cloud Core Services provide customer services such as service-to-service
integration, payment, billing services, identity management, search,
business process management services etc which can be consumed individually or on system to system integration basis.
Cloud computing provides several technological and business related advantages as presented in Fig.2.6. Cloud can provide different deployment
services models as described earlier. So in a way; cloud can provide anything
as a service including infrastructure, platform, software etc. New models like
Software + Services provide use rich client application hosted on cloud and
an interface to work with such application. Cloud provides different architectures as well as deployment service models and by using them its possible
to drive resource cost disruption and delivery innovation.
CHAPTER 2. OVERVIEW
19
Cloud provides several standard services [4] which are also known as core
cloud services mentioned in one of the cloud attributes earlier. These services
enhance the functionalities at customer end and provide tools to monitor
cloud usage, activities, performance, transaction etc. Different cloud services
and deployment models also provide differentiations as per customer needs
and they help in changing end customer behavior patterns.
Its important for any platform to generate complementary products and
services to achieve network externality and positive feedback [33] in the market. Cloud has ability to be flexible because of the virtualization and the
storage platform it provides. Along with different service delivery models,
its possible to innovate different services and products [18] by using cloud
platform. Also the economy of scale is easily achievable due to the fact
that cloud provides an illusion of infinite hardware resources for processing
complex tasks and storing large amount of data.
New business models like flat rate pricing, pay-per-use can be established
successfully using cloud architecture. Also its possible to cut the CAPEX
completely or partly for any company by using cloud services cleverly and
along with it, cloud drives reduction in the OPEX sharing computing resources with others [63]. This way cloud helps future business architectures
in transition.
2.4
CHAPTER 2. OVERVIEW
20
S models mentioned in section 2.3 can be utilized here for sharing resources. Cloud also provides sharing possibilities which is useful to
consume ideal resources from other system for processing tasks or let
other systems use ideal resources. This is possible in case of telecom IT
systems including self-service portals, CRM, billing systems and other
individual IT resources.
Cost reduction: Cloud has business model of pay-per-use payments
which gives possibility to let machines go when they are out of requirement. Cloud also gives option to start with small amount of resources and increasing resource utilization further when there is a need.
This elastic attributes make cloud a strong candidate to reduce telecom
CAPEX and OPEX.
Cloud services to enterprises and consumers: Cloud is relatively new
business phenomena in the market which gives telecom operators a
possibility to innovate and introduce new cloud based services to the
consumer and enterprise market. Also telecom operators can using the
network, branding image and marketing channels to sale cloud based
offerings to the consumer market.
In the current scenario, Major Telcos are more focused on the providing
cloud computing services rather than implementing their systems on cloud
computing platforms [10]. The main reason behind this is the critical SLA requirements defined for telecom systems and services especially in OSS area.
Service Level Agreement or SLA is an important requirement as well as
parameter for telecom systems delivering various telecom carrier-grade services. SLA is a document that specifies the rules of legal contract between
subscribers and service providers also stating technical performance details.
In telecom industry, SLAs are critical and often regulated by government.
Telecom vendors and operators often claim that clouds cannot yet meet the
telecom Service Level Agreement (SLA) requirements [45] [26]. For that reason a cautious approach is recommended. According to that a three step
roadmap depending on the carrier grade requirements is proposed as shown
in Fig. 2.7. Firstly, operators can utilize clouds in their support systems,
then step into tactical systems and finally to strategic systems [66].
Telecom systems have functions and services that are being implemented
on cloud by several operators, service providers and software vendors. The
possible scenario of cloud implementation of telecom software and services
[45] is depicted in Fig. 2.8.
Cloud implementation for telecom systems vary as per the application
features and attributes. There are different patterns [35] to deploy telecom
CHAPTER 2. OVERVIEW
21
Figure 2.7: Cloud Implementation of Telco Systems by Carrier Grade Requirements, adapted from [45]
CHAPTER 2. OVERVIEW
22
3. Software-as-a-service (SaaS): Cloud is used to deploy service delivery platforms and enterprise applications thus providing software-asa-service with cost-benefits and economy of scale. A good example is
salesforce.com which provides online CRM service on cloud for enterprises.
Public and private both cloud architectures could be used along with the
mixture of both of them i.e. hybrid cloud. Telco and several other generic
APIs are used as connectors between cloud and Telco domain.
2.5
2.5.1
CHAPTER 2. OVERVIEW
23
CHAPTER 2. OVERVIEW
24
2.5.2
Finance
Network operators focus on maximizing their profits from the subscriber base
and relative ARPU (Average Revenue per User) [50]. The basic calculation
for network operator profit can be presented with the below equation [30] :
(2.1)
Where;
ARPU = Average Revenue per User includes roaming and termination
apart from the revenue billed from subscribers. To maximize profits,
network operators try to increase the usage of the services and offerings
by providing more and better products.
Subscribers = Number of subscribers including existing subscriber base
and target subscriber base for the period of profit calculation. To maximize profits, network operators try to maintain existing subscriber base
(churn) and acquire new subscribers by service or price differentiations.
CHAPTER 2. OVERVIEW
25
CHAPTER 2. OVERVIEW
26
1. CAPEX: CAPEX is a capital expenditure required by telecom operators to spend while building the new systems or infrastructure. Developing or migrating existing applications to the cloud requires spending
expenses for enhancements in applications. Cloud deployment and integration requires expenses for setting hybrid or private cloud infrastructure and integration among systems.
2. OPEX: When hybrid or public cloud comes into the picture, the expenditures shall be counted as operational expenditures. Cloud related
OPEX includes per unit charge for processing, per GB charge for storage and high bandwidth network charges.
3. Non-OPEX: Carbon taxes, VAT and other surcharges shall be taken
into account as non-OPEX charges and shall be counted in the cloud
cost calculations.
4. Opportunity Gain: By using cloud, there are benefits gained apart
from cost reduction as well. Shorter time-to-market [39] and optimum
use of hardware resource driven by cloud elasticity shall be taken as
opportunity gained via cloud.
Chapter 3
Action Research
3.1
Action research [42] is done for real situations and real organizations suggesting improvements at the end as a result. Action research is an important
method as it consists of actual practices, data and results. In action research
researchers and practitioners come together. Problem finding, action analysis
and result reflections are important in action research.
For information systems, action research serves as an important process
where current situation analysis will find problems. Action intervention will
prepare the solution while designing, developing and implementing it along
with reflective learning which analyzes the collected data and discuss lessons
learned from the analysis. Researchers take tasks such as analysis, interviews,
team support, observations, findings and result publishing. One another
important aspect regarding industry or area specific action research is that
research helps industry and institutions to understand the problem situation
while tackling down the resolution for the problem.
The role of action researcher is very important for practicality of the
research. Researcher should be neutral to the situation while collecting as
well as analyzing details and data related to the situation. Researcher should
take a look at both positive and negative aspects of the data or evidence for
the situation. Researcher should set himself free of any perception or biased
opinions he has regarding the situation. Also sensitivity is another aspect
to be taken care of by the researchers to respond to delicate distinctions of
data.
27
3.1.1
28
To setup the action research project, there are certain steps need to followed
in order to get correct results. These steps are defined as following:
Problem Diagnosis: Find a practically relevant problem which also has
research potential. Obtain a general and comprehensive understanding
of the topic.
Action Intervention: Develop and define action research methods and
the research setup. Conduct the research and analyze the results.
Reflective Learning: Derive final results while gathering them from all
sources of research evidences. Build the report and documents representing the research objectives.
3.1.2
29
3.1.3
Interviews
Interviews [62] are considered as one of the very effective method of qualitative data research where expert opinions and practical knowledge of the
interviewees help forming the qualitative results. The basic strategy for the
interview is to gather multiple sources of evidences, create case study like
database and maintain chain of evidences. During the qualitative data analysis involving interview as primitive method of analysis, there are certain
points to be considered. Planning for the interviews is the first step in which
the data collection techniques should be linked to research questions. Goals
for the interview analysis should be defined afterwards keeping interviews
within the scope of the topic and not forgetting the sensitivity. Formulating
30
3.1.4
Delphi Method
Delphi method [49] is an iterative process to collect and filer out the anonymous opinions from the group of experts or customers or researchers using
the data collection tools and techniques interwoven with the feedbacks and
intermediate result analysis.
The goal of Delphi research is to improve understanding of the problems,
processes, future prospects, opportunities, solutions, ideas or to develop forecasts. Classic Delphi method varies by the group size, question formats, data
collections processes as well as tools and techniques involved in the research
[47].
Some of the important aspects of Delphi method are described as below:
Anonymity of interviewees or participants
Iteration and feedback mechanism
Result aggregation of the responses received
The basic structure of classic Delphi method is described in Fig. 3.1:
The classic Delphi method involves steps [49] described below in brief:
1. Develop the Research Question
2. Design the Research
31
3. Research Sample
4. Develop Delphi Round One Questionnaire
5. Release and Analyze Round One Questionnaire
6. Develop Round N Questionnaire
7. Release and Analyze Round N Questionnaire
8. Verify, Generalize and Document Research Results
Based upon the above described method the expert interview setup method
for the action research in this thesis has been derived. In the exit interview
setup for this research, two rounds have been considered following the steps
described in this section. The exit interview setup and also the steps for the
Delphi method involved in the setup are described in detail in next section.
3.2
The expert interview setup is based on the Delphi method described in the
last section. This setup includes two interview rounds of interview where first
32
round is done with the broader questions and scenarios while in the second
round questions are narrowed down with research specific data gathering
scenarios only. Fig. 3.2 describes the process flow of the expert interview
setup derived from the Delphi method.
1. Develop the Research Question: The research question was derived
in this research with help of the instructor and supervisor. Instructors
and researchers own industry experience helped with research problem
formation and inputs were also taken from the research program in
which thesis acts as part of the study. The literature reviews and pilot
studies done as part of the research program helped to formulate the
research background. Research questions were then derived from main
research problem and narrowed-down further to attain the research
targets. Research objectives were defined which denotes what research
goals will be achieved as part of the thesis research.
2. Design the Research: Research design is defined to plan the research
on micro level while reviewing different research methods taking their
pros and cons into consideration. The qualitative data analysis was
selected as a method of research with the expert interviews as a primary
technique of analysis for the thesis research. Research proposal was
then developed to formalize the research process and propose the idea
to the potential experts participating in the interview.
3. Research Sample: Selecting the participants for the Delphi process
is important as Delphi method considers the opinions of interviewees
as the expert opinions. Experts for the thesis research were selected
with focus on both technical and financial side of the telecom software
industry. The experts were from various telecom operator companies,
telecom software vendors, companies dealing with MVNO business process outsourcing and researchers from universities dealing in the MVNO
and telecom software domain. The research proposal was sent to the
experts for the interview considerations. Total 9 experts from 6 different companies and institutions considered the request and took part in
the research.
4. Develop and Conduct Delphi Round One: Delphi round one is
developed while doing literature review [66] [55] [29] with the broad
questions explaining the background and introduction of the research.
Round one questionnaire in the thesis research was semi-structured
which included open questions for discussions and structured questions
which include multiple-choice answers taking the reasons for the choice
33
34
8. Verify, Generalize and Document Results: The results were verified further with all the responses received from the experts in both
rounds of Delphi process. Results were then generalized where common results were taken together and differences were marked and documented as well. Further analysis was done and documented in following
section based upon the interview results.
3.3
3.3.1
Interview Responses
MVNO Systems
Question:
How does MVNO system work? (The main focus is on Business Support
Systems). What are the integral parts of the system? Please explain the
architecture in brief. Also explain MVNO business and operations in brief.
Response:
When setting up the MVNO, the first question need to be answered at
initial stage is why to setup MVNO? It generally takes 50-150 million euros
to setup a small to medium size MVNO. MVNO start-up decisions are made
based on the calculations of ARPU it will receive and no of subscribers in
nearly 2 years (called break-even point). Most MVNOs target to capture the
market as fast as possible and then make an exit (to be sold out). MVNOs
mainly work on service and price differentiation where time to market is
important due to fierce competition in the market. MVNOs should define
at initial stage whether they are a Telco, Media company or Media partner.
Some MVNOs also work as mobile advertisements aggregators.
Fundamentally MVNOs operate exactly the same as MNOs. Most operators have split their internal structure in three parts:
1. Customer relations and services, branding, marketing
2. Service layer includes product management, pricing, service management
3. Network, infra management, radio, network management
MNOs possess BSS systems and also network related systems i.e. OSS
and they have their own pricing structures for the products and services they
offer. In traditional MNO, IT systems are very complicated called legacy jungle but MVNO do not have such complex systems and also integration burden to replace the systems. MVNOs want to have bare minimum interface.
They buy network (the lower layer) from MNO or MVNE and implement
35
36
MNO and other partners. MVNO collaborate among partners to whom certain tasks have been outsourced. Service integration is also very important
for MVNO especially integration with MNOs and other partners in business.
MVNO need to choose the focus group or target customers while moving towards smaller segments of the market. They need to be much more
focused but the question is whether to focus on retail or enterprise sales.
Differentiation strategy creates base for MVNO strategy. Demographic, service, product, pricing differentiations can be adopted by MVNO. For example, an MVNO can target Russian speaking callers in Finland by providing
inexpensive calls to Russia. Some MVNOs also target niche markets of telecom services. For example, TeliaSonera targeted corporate customers during
initial stage with high quality of services by charging premium prices to customers. MVNOs focus mainly on post-paid residential market, so subscriber
hierarchy is simple. They need to focus on only one market at a time or
at a place. MVNOs can provide bundle solutions along with voice and data
services. In case of bundle solutions partner management is also important
for MVNO. Price differentiation is easy for MVNO but service differentiation
is limited due to dependency of OSS and network system on MNO. Most of
the MVNOs are price sensitive and keep ARPU at low level as to do effective
price differentiation. They have low cost and agile operations in place to
37
reduce OPEX. MVNOs work on price differentiation and tend to have happy
customers along with higher KPI. They have their own pricing structures for
the products. Subscribers win due to lower prcing mechanism but telecom
industry suffers because of high system administration cost. Telco industry
is a challenging business to be in due to fierce competition and so MVNOs
always look for additional revenue streams.
In recent years, it has been observed that MVNOs outsource as much
as possible because their operations are not about system but they are all
about the business. MVNOs dont care about system architecture and its
implementation. They just care about quality, branding, cost-benefits, SLA
agreements and customer satisfaction as well as retention. MVNOs use Business process outsourcing. Business process outsourcing in telecom domain
provides BSS as a service but its not Software-as-a-Service as there are customizations done for every telecom operator systems according to customer
needs.
MVNOs need lean business operations, business process outsourcing and
very effective as well as efficient systems and operations. MVNOs follow
rapid service delivery, development and marketing and so they need very
agile framework to support it. For MVNO to work there is no need to have
own processes, large organization and just by outsourcing it is possible to run
MVNO. Outsourcing is possible even outside the country but data security
and legal issues are critical.
MVNO do not invest on network so capital spending is low and they rent
network on pay-per-use or fixed monthly fees. CAPEX is low for MVNO.
They invest to get customer base and volume in the market and define threshold to acquire customer base. MVNOs want to have good discount as to get
volumes. They are getting the network capacity from already existing players so its regulated and competition is fierce for MVNO. MVNO pay some
fixed fees and also fees per subscriber. Subscriber churn is high in European
market around 20% and that makes churn very expensive for operators. For
MVNO, more you are dependent on MNO more challenges you have on differentiation side. Since MVNOs introduce competition in market, MNOs
dont want to give network capacity to MVNOs very easily.
Telecom network infra providers (MNO) rent the radio spectrum to other
MVNOs with high rates as they will face competition from those MVNOs
in future. Telecom network infra providers have lower rates for their own
operators, service providers.
Table 3.1 presents a situation in past with finnish telecom operator industry. There are several MVNO examples given below which were discussed
during the interviews:
Sonera Networks
TeliaSonera
Elisa
Elisa
Finnet Verkot
DNA Finland
38
Weak MVNO
Globetel
Finnetcom
NetFonet
Tele Finland
CDF Mobile
Cubio
Kolumbus
TDC Song
Fujitsu Services
GoMobile
Wireless Maingate
39
40
41
strong data encryption and consent from the operator. B2B customers do
not generally allow their customer information outside the country. MVNOs
in most of the cases have B2B private customers. Cross-location architecture
has to be in EU depending on the country regulations. Some MVNO may
not allow data going outside of country. Languages could be an issue, Bills
and online system in different languages
Cost-benefits:
Cost-benefits are heart of any business. Cost benefits the most important
thing or Cost factors are much more important for MVNO than other factors. Telecom business is a volume based business where price competitions
are fierce and low prices decide the volume. More cheaply you produce more
freedom you have upon pricing strategies. Outsourcing the systems is the
best option for MVNOs which are starting up their operations. SLA monitoring and ARPU calculations are also important besides the core business.
80% of MVNO cost incur to buy network capacity while marketing, admin
and resellers cost varies as per the subscriber base and MVNO strategy.
Qvantel BSS costs just 1% of MVNO spending. They provide BSS services on pay-as-you-grow basis just like utility computing. The business
model is based on the size of MVNO subscriber base. If MVNOs have fewer
subscribers then they pay less and when MVNOs grow with more number
of subscribers then they pay more. So outsourced BSS cost for MVNO is a
variable cost element and Qvantel share MVNO success as well as risks with
them. Qvantel asks a small percentage of ARPU for handling BSS operations
at initial stage and gets larger amount if customer base gets larger.
Along with these parameters, following parameters are important as well
for MVNOs.
Regulation aspects
Customer experience
Service type
3.3.2
MVNO on Cloud
Attractive sides of cloud are cost-benefits and economy of scale. BSS is easy
way to build, add functionalities in the system as BSS is easily configurable.
In-premises applications may need dedicated systems but outsourced systems
can use cloud easily. Several responses suggested that MVNO should put
all the systems that can be implemented on cloud. MVNO should be a
light weight organization and cloud can help a lot along with providing costbenefits and lean business operations. Apart from BSS, Revenue assurance
42
MVNO Systems
Self-service
(end-user facing
web-portal)
OMS
(Order
Management
System)
Integration
43
Processing
Bills
Mediation (processing
call
records)
44
Bill Formatter
(Formatting
and
printing
bills)
Collecting
Payments
Account
Receivable (receiving payments)
Collection (due
payments, barring, processing
payments)
45
Rating has also problems similar to Mediation. Invoicing can be deployed on cloud easily. Qvantel
is already doing it. Small operators tend to use it.
Network bandwidth should be large and hypervisors should be fast. There are latency driven issues
as well as dependencies on IO and hypervisors.
There already some cloud implementations of bill
formatter in place and some companies are already
providing it. For example, StreamServe provides
bill formatting and printing as a service. There
can be problems as bill formatter involves physical
systems such as print devices and proper networking is needed in such cases.
There is not a major problem to implement account receivable on cloud however there are concerns regarding credit card information security.
Data access and security, especially for financial
data, must be of high level for this system. Number of processes are less so account receivable may
not have significant advantage if they are implemented on the cloud independently.
There is not a major problem to implement collection on cloud however data security and data
access could be an issue and cloud implementation
of collection system depends on the service layer
implementation. Collection has high amount of
processes which gives possibility to implement this
system on cloud if other parts of BSS are implemented on the cloud. TeliaSonera has it on cloud.
TeliaSonera has outsourced it to collection agencies and they are running the system on cloud.
Internal collection processes, network barring and
debt collection agency processes may get affected
and need attention.
Inventory
Prepaid Operation
Prepaid Card
System
(real
time charging)
46
In many cases, MVNO rents inventory management system from MNO. If a service layer is in the
cloud then its easy to deploy it on cloud as ERP
/ CRM will already be on cloud making inventory a viable candidate for cloud. It would follow
same SLA as in case of self-service. However, the
counter arguments suggested that the inventory
system is close to network elements such as provisioning and others which make it difficult to deploy
on cloud due to regulatory issues. For MVNO,
it might be difficult to implement all systems as
cloud may act as generic interface for different elements. There is only advantage of cost sharing
but it might be the strong reason to proceed with
implementation.
The common consent from interviewees was that
prepaid system is challenging and not feasible from
technology point of view to implement on cloud.
The reasons are described as following. Cloud implementation of prepaid systems is difficult as its
a real time charging system and there high availability requirements for such systems. SLAs are
critically high and no delays should be there. Intelligent network systems are very close to prepaid systems and high level integration is needed
which is critical. It will be easier with flat IP architecture in future but right now implementation
is way too risky. However, there were some positive responses for cloud implementation of prepaid
system which mentioned that online top-ups can
be easier with cloud implementation however real
time charging setup can be challenging. There are
real time charging systems from TeliaSonera for
pre and post paid which can be deployed on cloud
easily. Country specific clouds are good options to
implement prepaid like systems.
47
Question:
If following parameters are affected while migrating some part or the whole
of the MVNO system to the cloud? How?
Response:
Parameter
Performance
(Throughput,
Latency)
Service
Level
Agreement
(SLA)
Service Delivery
Customer
vices
Ser-
Security (N/w
and Data Security)
How is it affected?
Performance is important and MVNO would like have the same
performance level whether or not the systems are implemented on
cloud. Throughput and latency are affected on cloud but network
bandwidth upgrade could save the situation. Tight data center
SLA could improve the performance.
On BSS side, SLA requirements are not critical especially for
retail or residential customers. SLA may not have critical impact
on Billing and CRM systems but the monitoring is needed. For
corporate customers, billing SLAs are important and may get
affected. Also, internal SLAs between MNO and MVNO may
have issues. Cloud service provider should take care of SLA and
availability factors. Some responses mentioned that SLA depends
on implementation and architecture of IT systems.
Service delivery is not affected much and it may not be visible to
customers that the system is implemented on the cloud. In case
of self-service portal migration to cloud there is a possibility that
service delivery may get affected. However, most of the interviewees agreed that sufficient self-service can be organized with the
cloud.
Customer service is important and critical for MVNOs. It gets
affected if system implementation on cloud is not proper and this
may ruin MVNO business. Costumer services are mostly implemented as self-service portals (or CRM) and it highly depends
on technology used to build the system. However, some interviewees agreed that customer services wont be affected much. At
present many small companies are using CRM on cloud solutions
and they are getting good benefits without significant problems.
Security of Call data records and customer sensitive data is critical
and the most important factor to consider while moving MVNO
systems to the cloud. Data security and availability are highly
affected. Recent issue with Sony Play Station cloud is one of the
examples. Security is easy to implement if systems are in-house
but cloud implementation has several challenges.
Carbon
prints
Foot-
Business Model
changes
Cross-location
Architecture
3.3.3
48
Cost-benefits
Question:
Which cost factors are involved in telecom systems especially for MVNOs?
How do you calculate pricing for telecom cloud? How do you get Return on
Investment (ROI) while implementing MVNO systems on the cloud?
Response:
MVNO prices are very much competition driven and not cost calculation
driven however cost-benefits can be driven by moving implementing MVNO
systems on cloud which makes MVNO further lower the prices. Cloud brings
advantages for MVNOs to cut system administration costs while moving from
49
50
51
can help. Correlation for price competition among operators. Oligopoly settings among operators. Along with cost-benefits; assurance, legal aspects
and flexibility factors are also important for MVNOs as to adjust themselves
with the competition as fast as possible.
3.3.4
Question:
What are the implications of cloud deployment of MVNO systems? How
government laws can affect the deployment? What are the financial implications?
Response:
Regulatory restrictions from government related to the customer data
security and transferring data outside the country are the most discussed
implications in the interview sessions. Data security and data access of customer sensitive data are the biggest challenges. In some countries, there is
a law that authorities can listen to the communication lines. Moving data
outside of country may not be possible as per some customer agreements.
For example, police and fire-brigade numbers cannot be taken outside the
country. If such important data need to be put in cloud then highest level
of security and availability should be in place. However, bilateral contracts
done between or among countries may help lowering the effect of such implications. One point of view related to these implications was regarding
customer data that its possible to move CDRs among EU states with consent from the customers. Its interesting to note that there were no major
implications faced by Qvantel while moving MVNO/MNO BSS systems to
the cloud.
There is also a challenge that the customer data should be retained in
BSS systems for specified time frame and it should also be possible to remove
data which can be problematic with the cloud implementation. Its very
dangerous for operators if customer data is stolen or lost. For example, Sony
PlayStation PS3 customer credit card data were stolen. Such situations may
put operators in difficult position and operators could lose the significant
customer base and business. Operators must have access to customer data all
the time. Union or government regulations are general implications. Security,
privacy and network could be technical implications. Security while sharing
processing and storage spaces with other companies is a big concern.
There are problems with ability to do customization, systems integration
and configuration challenges. MVNOs may want to change BSS functions
at times and cloud makes it difficult. Time to respond to new legislation
changes could have significant delays. There are also regulations from tele-
52
com authority related to data, SLAs and customer services which should be
taken care of while moving MVNO systems to the cloud. On the operational
perspective, the issue is to maintain critical tickets in the region as to solve
customer issues within the region which may be difficult if data is not available with operators all the time. Interaction transaction logging is also a
challenge especially in case of B2B customers if systems are on cloud.
Latency and throughput problems may arise if services are migrated or
outsourced to other country. Country specific requirements or factors such as
call loop timings, data protection regulations, network latency etc may create implications. Labor laws are also of concern as operators may reduce the
workforce from IT admin department to gain cost-benefits from cloud implementation. Generally 20-30 human resources reside in IT admin departments
in medium scale MVNOs and implementation of cloud gives possibility to reduce IT admin resources significantly. No financial implications as such.
Question:
What do you think of the future of MVNO and Telecom Software industry
also considering importance of cloud? What vision do you look for?
Response:
Cloud in Telco IT systems is already taking place. However, Telco clouds
with high availability and closing the gaps of carrier-grade SLAs are not yet
available. OSS on cloud is not feasible as of now due to high OSS SLAs,
network bandwidth problems, and cloud issues of availability and SLA situations. However, Comptel is planning implementation of several components
of OSS on the cloud. So its clear that for Telco, there is still long way to go
to deal with cloud fully. There are inherent problems in network architecture
and so MSS and HLR on cloud are not feasible.
Moving existing systems to the cloud is more difficult than starting the
operations from scratch on cloud. Cloud is important where capacity in network or processing might not be enough. Operators ted to get market share
and cloud will help with its economy of scale attributes. There will be future for the cloud when its technically feasible, economically cost-beneficial.
However, cloud is fragile as lot of network related and trust issues to have
data outside company or country or region.
Future will be more like utility based telecom traffic and implementation
of flat rate business models. Flat rate data plans are already in place and
getting popular. There will be no need of complex billing systems for flat
rate pricings and such scenario will work in favor of cloud implementation
of telecom systems. Clouds will be very important in the future as they are
also based on utility computing architecture. Also Cloud will be a natural
choice for MVNO in fierce competition of price and service differentiation.
If integration and data security is proper then Telco will tend to use cloud
53
services and gain benefits of cloud making cloud - a good, smart and reasonable hype. BSS and CRM will be easy to deploy on cloud however OSS
and Prepaid wont be easy to move to the cloud. In case of MNOs, lot of
MNOs have old in-house legacy IT systems and its difficult and expensive to
move or migrate those systems to the cloud. TeliaSonera is using salesforce
in Denmark and Norway and they are also evaluating possibility to expand
it more. However, TeliaSonera is not aggressive on cloud front.
3.3.5
Round 2 Results
Sr
No
2
3
4
Statements
Percentage Response
(single choice)
Not
TotaSure
AgrDislly
or
agree
Agr- ee
Deee
pend
Totally
Disagree
100
100
60
40
20
60
20
5
6
7
10
54
40
40
20
20
60
40
20
20
The results are analyzed further in section 4.1 in which the graph is
presented with the response analysis of each statement and its importance
in context of MVNO systems implementation on cloud.
Chapter 4
Techno-economic Analysis
4.1
4.1.1
The first round of interview had first two questions related to MVNO systems without introducing cloud to get idea how experts think MVNO systems work. The responses obtained were ranging from the technical and
operational to business perspectives. The main points from all the responses
combined are given in Fig. 4.1.
The responses presented in Fig.4.2 are collected during the first round
of interviews in which the structured question-answers were asked to the
interviewees for mapping the MVNO systems on cloud. There were three
options for interviewees to choose from and those were 1. Agree, 2. Not
sure or depends and 3. Disagree. The graph below describes the responses
in an analytical manner in which the MVNO systems are mentioned on xaxis while y-axis refers to the percentage response obtained in reply to the
question whether system should deployed on cloud?
By looking at the graph in Fig.4.2, we can say that the experts were quite
insistent to put self-service, CRM and bill formatter systems on cloud as to
provide SaaS like environment for such applications. For Order management,
Billing & CRM integrator and account receivable systems the responses had
some doubts and so small portion of responses were not sure to deploy such
systems on cloud. The experts thought of integration problems with other
systems be the main issue in these cases. Rating and invoicing had similar
responses in which close to half of them were not sure to move these systems
to cloud. The interviewees mentioned data transfer as well as latency driven
issues with cloud I/O and hypervisor [25] be the key concern in these cases.
Its interesting to note that all of the above mentioned systems had positive
55
56
or neutral answers which make them viable candidates for cloud after solving
key issues associated with them.
Product catalog, collection and resource management had overall positive responses however there were some neutral and negative reactions at the
same time. Integration with external systems as well as proximity to the network elements drove such responses for these systems. Mediation remained a
doubtful system for the possible cloud implementation as it got same number
of positive and negative answers along with half of the responses as neutral
or uncertain. This was because of the CDR security and transfer problems in
the cloud environments. Prepaid as well as network systems got more than
half of negative responses and so they remain impractical to implement on
cloud as of now. This is due to a well-known fact that cloud is not ready for
telecom systems to provide carrier-grade services [38].
Fig. 4.3 illustrates the same interview responses in a different manner in
which it categorizes each MVNO system according to the percentage of positive (Agree) response acquired in reply to the question whether system
should deployed on cloud? This figure depicts almost the same picture as
mentioned in analytical description of the graph. Systems with green should
be moved to the cloud while systems in blue can be moved to the cloud after
resolving related issues with cloud or the systems themselves. Mediation is
the only system in violet and it remains suspicious for cloud implementation
while prepaid, OSS and other network systems remain repulsive to move to
the cloud.
MVNO systems drive several advantages which were discussed during
the interview sessions with experts. The focus remained on the system performance and cost-benefits driven by implementation of MVNO systems on
cloud. Fig. 4.4 illustrates these advantages in brief.
57
4.1.2
Application Mapping
58
As described in section 2.5.1 the application mapping for MVNO business support systems [34] can be done by mapping application attributes to
various cloud scenarios. Table 4.1 maps two to three main attributes to the
possible cloud scenario for each application [28] and then decides whether
the application should be deployed on cloud or not.
MVNO
Operation
MVNO
Systems
Taking
Orders
Self-service
Portal
Key Attributes
Cloud Scenario
User experience
Scalability
Availability
SaaS
Economy of scale
99% availability
Deployable?
Yes
CRM
Integration
Transaction/data
Order
locking
Management
Complex integration & customization
Load balancing
Billing &
Transaction manCRM
agement & persisIntegrator
tency
Scalability
Product
Catalog
Processing
Bills
Depends
Hybrid cloud
Unpredictable performance
Configuration management
Online data
Configuration
on cloud
SaaS
Depends
Data transfer
Data security
Bill
Formatter
Critical
Integration
obstacles
Yes
Depends
Batch processing
Invoicing
Elasticity on cloud
Mediation
Rating
59
Configuration management
Batch processing
Offline data
Scalability
Batch
processing
(computeintensive)
Monitoring & affordability
Physical
device
data transfer
issues
No
Depends
Yes
Collecting
Payments
Account
Receivable
Collection
Inventory
Prepaid
Operation
Batch processing
Efficient
processing
Online data
Data security
SaaS
Data security issues
on cloud
99% availability
Data lock-in and security problems on cloud
Integration obstacles
Efficient
batchprocessing
SaaS
Integration & configuration issues on cloud
Cost-beneficial implementation on cloud
99% availability - not
feasible
Performance unpredictability
Integration challenges
Availability
Data locking & security
Integration
Batch processing
Online data
Resource
Integration / conManagement
figuration
Cost-sharing
Real-time
Charging
60
High
availability
(carrier-grade)
High performance
batch-
Integration
Table 4.1: Analysis for mapping MVNO systems attributes
on cloud scenarios
4.1.3
Strategy
Yes
Yes
Depends
Yes
No
61
good data encryption policies is a good option especially for the MVNO
systems. Applications with parallel batch processing, analytics or business
intelligence gathering and computation-intensive tasks should be moved to
the cloud. The customer-facing systems such as self-service portals and CRM
should be deployed on the cloud while providing SaaS to end-users.
Business:
Need to check regulatory requirements from authorities. Data availability and retention plans as per regulations should be in place for customer
sensitive data. Cost should be calculated for moving the systems to cloud.
This includes cost for application migration, customization and integration
changes as well as monthly cost for processing, memory and network bandwidth usage cost for cloud. Its good to have in-country or in-region cloud
instances in case of public or hybrid cloud deployments. Its important to
find as well as mitigate impacts on SLA and customer services while deploying customer facing systems on cloud. There is an option of business process
outsourcing or manages service outsourcing to the vendors providing such
services for more cost-beneficial approach.
4.1.4
Proof of concept:
The public cloud involves the data and application access via public internet channels and so many critical services and sensitive data might be
exposed to the security threats on the public networks. The private cloud
is more restricted environment which resides internally or nearby company
premises constituting safer environment for enterprise applications and re-
62
lated data. However, private cloud may not significantly represent the benefits of using cloud such as elasticity, economy of scale, cost factors, green
computing etc. Hybrid cloud is the model which will prevail [22] in such
cases where its possible to install database only in private cloud and application images residing in both private and public cloud will access it through
highly secure data channels. This way an enterprise can make efficient use
of internal IT resources [24] and they can switch dynamically to the public
resources whenever there is a need. Following are the steps derived to implement enterprise applications on hybrid cloud environment (See Fig. 4.5)
while balancing the load among private and public cloud instances:
1. Setup a private cloud and create private cloud instances. The private
cloud instances are the chunks of the whole private cloud resources
combined and it should be designed in such a way so that optimum
memory and processing power are allocated to each chunk or instance.
2. Install application database (or databases) and build an application
image for different applications on private cloud. Ensure that all the
application images of different applications have necessary configurations and database access on private cloud. Application images are
virtual and ready-to-use installation of applications.
63
3. Modify load balancing script on underlying server in use so that application can start using cloud instances to balance the incoming load.
4. Setup the public cloud instances and get elastic IPs of those instances.
It should be noted that the use of public cloud instances from multiple
vendors ensures high availability as in case of down-time of cloud instances; system can start using cloud instances from other public cloud
vendors.
5. Add public cloud instances (IPs) to load balancing script. Along with
this, also add the logic to switch from private to public cloud instances
whenever the private cloud instances reach at their maximum usage in
terms of memory or processing or other attributes depending on the
application purpose. There are different technologies and tools [52]
available which can help creating such environment where both private and public computing resources can be used efficiently by defining
underlying requirements.
6. Allow application and database to have external access with encrypted
channels. The channel encryption shall be supported by highly efficient
network security algorithms in place. Virtual private network and high
bandwidth data pipe can be established in such cases.
7. Prepare auto-spawning script for public cloud instances so that when
system switches the processing from private to public cloud, the public
cloud instances will be started automatically. Similarly, add logic to
turn off public cloud instance where the processing need is ended on
that instance. This functionality is needed for the reason that public
cloud usage is charged on pay-per-use basis and auto-spawning reduces
the cost and adds to the dynamic elasticity of the system resources.
8. Add monitoring and logging scripts to track the system usage and
operations. It should be noted that web-based system can be accessed
by single IP of a load balancer.
Experiment:
In the experiment (See Fig. 4.6) followed from the proof of concept presented above, a private cloud was setup on virtualization platoform and private cloud instances were created. After that, the installation of sugarCRM
mysql database was carried out and application image for sugarCRM was
compiled. Already built sugarCRM images were then deployed on different
private cloud instances ensuring that all the application images of sugarCRM
had necessary configurations and database access on private cloud. An http
64
65
jobs. Also, auto-spawning script for public cloud instances could be prepared
so that when system switches the processing from private to public cloud,
the public cloud instances will be started or turned off automatically. The
web-based sugarCRM was accessible by single IP of an http load balancer.
4.1.5
Analysis of Round 2
As stated in the round 2 results in section 3.3.5, round 2 was developed while
narrowing down the questions or scenario while addressing the important
statements created after analyzing the responses from round 1. The graph in
Fig. 4.7 illustrates the statement wise single choice responses acquired from
interviewees while statement wise brief analysis in done as following.
1. MVNOs implement or outsource BSS and customer care systems. They
rent the radio capacity from network operators and use their OSS.
Analysis: This statement got all the responses in total agreement and
so we can say that while analyzing MVNO systems the OSS and network systems doest come into the picture directly. Therefore, MVNO
systems generally do not possess carrier-grade SLA and their implementation on cloud is possible without Telco carrier-grade services.
66
67
4.2
Cost-Benefits
Cloud may not look that attractive while comparing the server utilization,
memory requirements and data transfer of small or medium sized in-house IT
systems with cloud side by side. However, several additional cost savings such
as power, cooling cost, physical plant space, IT administration cost, insurance
etc [5] makes the cloud more interesting in case of cost-benefits scenario. The
main idea of moving from CAPEX to OPEX while further reducing OPEX
may not save a lot of money for the company but its definitely an attractive
option for companies starting up their operations. This transition minimizes
the risk on investment for a new entrant.
Another aspect is to reduce huge operational and administrative cost of
the company by using cloud. One statistics says that often companies spend
8$ for administration of hardware cost of every 1$ [59]. By using cloud, this
huge gap between hardware cost and its administration cost can be eradicated considerably. Cloud provides on-demand utility based computing in
which due to low capital expenditure the total cost of owner ship (TCO) and
lower profitability over the period. On ARPU, however, cloud has positive
impact and cloud can drive increment in ARPU and reduction in subscriber
churn. Equation 2.1 shows that as CAPEX and OPEX decreases the profits
increases.
68
MVNOs are the early adopters of SaaS based solutions and its the same in
case OSS/BSS too. MVNOs always look for rapid deployment of the system
as to have faster time to market. There is also a risk involved in MVNO
operations of overestimating investments cost and underestimating demands
which may create bad reputation for the MVNO. Cloud helps significantly
while almost eliminating such risks [65]. It lowers the initial investment while
reducing barrier to the new entrants in market and so MVNOs will get high
revenues if successful and low revenue losses of not successful.
The Fig. 4.8 presents the main points discussed during the expert interview sessions in the area of cost-benefits while moving MVNO systems to the
cloud.
4.2.1
Cost Calculations
Table 4.2 presents an example cost calculation for CDR processing on mediation platform by taking following parameters and their nearly practical
values received from interview responses or literature reviews:
No of Subscribers: 10000 to 1000000 for typical MVNOs
ARPU: 20 Euro (in Europe and Americas) while 5 Euro (in APAC)
CDRs: 200 CDRs/subscriber/month 7 CDRs/subscriber/day
Processing speed: 2000 CDRs/sec
Memory: 1 KB/CDR
Network transfer speed: 20 Mbps
No
of
subscribers
No
of
CDRs /
day
10000
70000
Memory
(in
GB) /
day
0,07
50000
350000
0,35
100000
700000
0,7
500000
3500000
3,5
1000000
7000000
5000000
35000000
35
69
By looking at the data presented in Table 4.2, it shall be noted that till
1000000 subscribers its feasible to use the public cloud packages. However,
after 1000000 subscrbiers. it gets difficult to use only the public as the data
transfer costs increase heavily and in this case hybrid cloud is a good option
to implement the CDR processing platform.
Another example of cost calculation [31] can be given for the SugarCRM
implementation on hybrid cloud setup presented in section 4.1.4.
Server or instance configuration: Private cloud server: Intel e5640 or i5
dual core or equivalent, 2.6 GHz, 4 GB RAM, 8-12 MB Cache Public cloud
instance: 2.6 GHz, 4 GB RAM
How to calculate private cloud cost [23] for 1 server over 3 years: (1
server * 1500 $/server) + (3 * 8 * 1500 $) admin cost/server + (3 * 800 $)
electricity cost/server = 39900 $ Note:Admin cost is taken into calculation
with the fact that companies spend 8 $ of administration cost for each 1 $
spent when buying the server.
70
How to calculate private cloud cost [23] for 1 instance over 3 years: ((1
instance * 0.50 $/hour) * (3 * 365 * 24) hours) + 1500 $/year network cost =
14640 $ Note: 0.5 $/instance can be broken up into 0.2$/hour computation
cost + 0.15 $/GB data transfer cost + 0.15 $/GB data storage cost (total
storage cost divided)
Cost calculations for SugarCRM [54] on hybrid cloud over 3 years are
presented in Table 4.3.
Users
1 to 5
5 to 50
50
100
to
Max
Server CPU
Records
1,000,000 1x i5 core or equivalent
2,000,000 2x i5 core or equivalent
4,000,000 4x i5 core or equivalent
Server
RAM
4 GB
4 GB
8-16 GB
Cloud Usage
1 private instance
2 private instances
2 private &
2 public instances
Cost
39900 $
39900 $
39900 $ +
29280 $ =
69180 $
Table 4.4 shows cost comparisons for SugarCRM on private, public and
hybrid cloud over 3 years.
Users
1 to 5
5 to 50
50
to
100
Max
Records
1,000,000
2,000,000
4,000,000
Cost
brid
39900
39900
69180
for hy$
$
$
4.2.2
Opportunity Gains
MVNOs can also gain siginificant advantages by using cloud apart from costbenefits and such advantages shall be taken into consideration as opportunity
71
gains. One of the important gains is the faster delivery of customer services
while reducing the time-to-market [39] parameter. Cloud supports the webbased services very efficiently and companies can provide the applications
and services on SaaS based cloud solutions quite effectively after making
business decisions.
Another aspect is the green IT and evironment friendly computing. Cloud
saves the power consumption and enxpenditure on hardware resources along
with reducing hardware cooling needs significantly. There are several MVNOs
already marketing their products with green initiatives. For example, an
MVNO Ecofoon in Netherlands. Cloud will help adding another aspect to
the green initiative by acting as an environment-friendly hosting platform.
There are also ideas of green cloud [3] coming into the picture to enhance the
efficiency of telecom operators. Along with above impacts, large MVNOs can
expand their service offerings to the end users as well by introducing cloud
services. Such cloud offerings can gain huge benefits to the operators without
adding too much costs to setup and market them.
4.3
4.3.1
Parameters Evaluation
Importance in MVNO Context
Its important to see that in MVNO if scenarios such as performance, carriergrade SLA, cost-benefits and cross-location architecture are important. To
map MVNO systems on the cloud its vital to know that how these parameters behave in MVNO context. Fig. 4.9 shows the graphical analysis of
the structured interview responses for the paramaters importance in MVNO
context. Performance [41] and carrier-grade SLA [45] are affected in case of
MVNO implementation on cloud and may not retain the same level as in
case of normal implementation without cloud. Cost-benefits are affected in
a positive manner while using the cloud and cost-benefits can alone drive
the possibility and decision to implement MVNO systems on cloud [55].
Cross-location architecture is important as cloud provides distributed and
centralized operations and therefore MVNOs operating across the countries
have significant impact on their operations.
As we can see in performance case, more than 40% responses agree that
performance could be an important scenario in MVNO context. As we see
in results, performance seems to be important in cases of network systems,
CDR transfer and prepaid operations (real time charging). Some responses
mentioned that the performance really depends on the MVNO systems, their
size as well as implementation. Small and medium MVNO systems do not
72
need high performing systems as such due to small amount of data traffic
in and out of the systems. However, large in-house systems do need high
performance standards as to handle the system load efficiently and drive
customer satisfaction to the highest level. Two of the responses mentioned
that MVNOs now tend to outsource or rent BSS. In such cases, MVNO
do not care about performance of the system but they care only about the
SLAs agreed among MVNO and other parties to which system operations
are outsourced.
The problem with the cloud not becoming Telco cloud which can offer
wholesome implementation of Telco software is the carrier-grade SLA requirement often defined by government. However, by looking at the responses we
can say that at least on the BSS side or for the MVNOs it doesnt matter
much. Carrier-grade SLA is often defined for OSS and network level systems
having SLA in degree of 99.999% or more. As mentioned in results, billing
and the front-end systems in BSS have SLAs at around 85% and 98% respectively. Also, public cloud provides SLA and availability close to 99% [53]. So
we can say that carrier-grade SLA may not have much to do with MVNO
BSS. Nevertheless, if prepaid systems are involved in the MVNO operations
or MVNO possess HLR and other switching elements then carrier-grade SLA
matters significantly.
From the results shown in the graph, cross location architecture is important in MVNO context as there was not a disagreement for the scenario. More
73
than half of the responses recommended that cross-located or distributed architecture is important while other responses suggested that it really depends
on the MVNO operations, business model and the market they are operating
in. Lot of MVNOs operate across the countries or continents in some cases
[12] and building cross-location distributed architecture brings benefits at
system as well as financial levels. Even for MVNOs which are operating in
one country only, web-based distributed CRM and order management systems are in place and cross-location distributed architecture helps to achieve
economy of scale as well as centralized cost-beneficial operations while utilizing hardware resources at its full capacity. There are problems in MVNO
cases operating in different countries or geographic locations where the government regulations restrict the transfer of customer sensitive data outside
the country or region [61]. Also systems like fraud management needs data
to be available all the time to track the requests from authorities. However, with customer consent and proper data encryption and access policies
it could be possible to transfer CDRs and other system data across the region
which helps in building cross-location architecture.
By looking at the responses for cost-benefits scenario, its clear that the
most important thing for MVNO is the cost-benefits across different areas
of operation. More than 85% of responses suggested that cost-benefits have
the significant importance in MVNO operations forming the core of MVNO
business. There are already sections in the thesis which mentioned about the
cost-benefits and its importance for MVNO. MVNOs compete on the basis of
price differentiation in most of the cases and cost-benefits driven from MVNO
operations can be utilized to drive the price competition further achieving
the edge in the market.
4.3.2
If MVNO systems are moved to the cloud then parameters which are important in MVNO systems context may get affected. These parameters can
impose positive or negative effects on MVNO systems and business [16]. Fig.
4.10 presents analysis of certain important mvno paramters in cloud settings
and their influence to the MVNO systems implementation on cloud. As
stated earlier in this section, performance is an important criterion in MVNO
systems context. 8 out of 9 responses believed that in one or the other way
performance will be affected in case if MVNO system is implemented on
cloud. This should be taken as an important point as in high-tech telecom
systems, performance along with throughput and latency matters significantly and cloud might not be ready for it yet. SLA has mixed responses
where around 55% responses mentioned that SLA gets affected but not crit-
74
75
Chapter 5
Discussion
5.1
Billing as a Service
Billing-as-a-service is to be setup on the hybrid cloud in form of platformas-a-service using the proof of concept presented in section 4.1. In this architecture, database remains in the private cloud due to data security and
access reasons. The real time and batch job processing is to be done on both
private and public cloud instances. To handle fluctuating load, hybrid cloud
provides possibility to setup private cloud utilization for average processing
needs. For the load going beyond the capacity of private cloud, public cloud
instances are to be spawned when needed. This way it will be cost-beneficial
to maintain the system and provide multi-tenant platform to different companies together. The public cloud instances will have access to the database
on the private cloud with highly encrypted network and cloud data security.
Load balancing and auto spawning of public cloud instances will ease the
administration of the system. Data security, access, transaction locking and
API integration are the areas to be taken care of.
Fig. 5.1 shows the proposed architecture at high level for billing-as-aservice implementation. Network usage data comes from MNO to the system
in form of CDR and IPDR via highly secure data transfer channel. Robust
mediation system converts the usage data from different format to a common
standard and places the data into the database. Rater provides interface to
rate the calls and data traffic for subscribers. It takes input related to the
products rates and offer changes from reference database. It also gets input
related to call durations, geographical destination of calls made, roaming,
time and origin of the call related to each call made by subscriber. It converts
the call record data to financial data and forwards it to the invoicing system
further. Here in this case, hybrid cloud makes a good case since mediation
76
CHAPTER 5. DISCUSSION
77
and rater load fluctuates as per the input coming from external systems. Its
possible to use necessary public cloud instances only for processing xDRs and
not storing the customer sensitive database.
Invoicing and billing system then analyzes and calculates one time and
recurring charges to the subscriber call usage according to the price plans
and offers selected by subscribers. It also calculates the discounts and taxes
for the subscribers as per the current government regulations and ongoing
marketing campaigns. After all calculations it computes the final bill amount
for the subscribers and prepares necessary data for the bills. Bill formatter
system gets all the data needed from billing system to print and format it
so that it can be printed or sent to the customers by e-mail services. Bill
distribution prints and sends bills to customers or post agencies. It connects
to external system or self-service portal to put electronic bills to subscribers
if needed.
Payment gateways like PayPal etc are connected to the billing-as-a-service
platform so that customers can pay bills via their credit cards or online
bank transfer. Such gateways can be integrated with self-service or CRM
platform. Account receivable connects to self-service or CRM and payment
gateway to collect the payments and dues from customers. It stores the data
of payments received along with reference no and other payment related
information. Collection collects subscribers payment information and sends
barring or unbarring instructions for the subscription to the network as per
the rules defined in the system. Collection may also have integration with
debt collecting agencies to push due payment collection.
Product catalog, CRM and self-service portal acts as external systems
to the billing-as-a-service platform and they are connected with APIs to the
billing service platform. Product catalog provide offers and services related
data to the system. CRM enters customer data like name, address, price plan,
offers, campaigns etc to the billing system and it gets billing data from billing
platform to display it to the customers or customer service representatives.
Self-service portal gets bill data or even electronic bills to present it to the
customers on the web service portal.
In future, billing systems may have flat architecture in place and it may
be possible to provide billing services without much customization and configuration changes [32] [19]. Prepaid and flat-rate billing-schemes are already
getting popular and for them complex billing systems are not needed as such.
Billing-as-a-service will be a feasible idea at least for small and medium scale
MVNOs to drive cost-benefits further down. The service should be multitenant where a single instance of the software runs on a server, serving multiple clients. The service can be designed on cloud to virtually partition its
data and configuration, and each client organization works with a customized
CHAPTER 5. DISCUSSION
78
CHAPTER 5. DISCUSSION
79
illegal access shall be possible to the data. Country specific data should
remain inside the country most of the time. Data security must be of high
level in any case along with which the security data backup and replications
must be planned.
5.2
Even though business models are evolving, we may not see any radical
changes in MVNO business models due to cloud implementation of their
systems. There may be incremental changes which are competition driven
or driven by the cost-benefits obtained by cloud deployment [16]. Telecom
software vendors and managed service providers will try to expand their role
by becoming true partner to have longer and stronger relationships with CSP
rather than just selling services. In this way, CSPs can focus on their core
business of selling communication services to the end-users and move responsibilities of their software systems to managed services partners. Cloud
can help to drive cost-benefits in this case since managed services providers
can support multiple CSPs simultaneously to allow efficient resource sharing
among partner CSPs easily. For example, before going to the cloud, Qvantel
was traditional software and a platform company. Now Qvantel is Business
Process Outsourcing partner or a managed services company for telecom service providers operating several MVNO platforms at the same time. They
take care of billing and collection runs, performance and optimization of
system etc by their own platform.
One of the main obstacles for MVNO to enter CSP market is the entrant
cost. MNOs see MVNO as potential threat to their business and increase in
incumbency factor [17]. So MNOs are always reluctant of sharing network
with virtual operators but because of the government regulation they must
have policy in place of sharing their network capacity among virtual operators. This drives very high network sharing cost for virtual operators as
MNOs try to maximize their profits by getting higher prices for the network
sharing. MVNOs may spend 70-80% of their revenue for buying the network
capacity from network operators. Cloud computing provides significant benefits for MVNOs where it reduces the fixed cost and operating costs as well
while implementing the business support systems and other integrated systems. It reduces the cost to enter the market for MVNOs and thus driving
business creation along with higher competition in the market.
Cloud provides very flexible processing and storage architecture due to
which its possible to achieve economy of scale [5] for network operators.
Low operational expenditure can be driven with cloud implementation and
CHAPTER 5. DISCUSSION
80
5.3
Implications
As mentioned in the interview results, there are still major implications [37]
involved with the cloud implementation especially of the enterprise systems.
Fig. 5.2 presents the snapshot of the implication findings obtained during
the expert interviews. The analysis of these implications is done as following:
Availability of service: Cloud has availability close to 99.9% which
might not be enough to support telecom operator business and due to availability issues; operators may lose business as a result of down time of the
systems on cloud. Telecom software implementations, especially on real-time
charging, fault tolerance and network switching side should have availability
close to 99.999% of the time and recovery time of around 5 minutes which is
50-100 times higher than normal public cloud availability [38]. There were
outages reported due to programming or software errors at cloud provider
companies like Amazon, Google, Microsoft etc and those outages were of durations ranging from 2-8 hours [5]. Such availability issues show that cloud is
not ready to host carrier-grade telecom services as telecom operators are very
sensitivity about the business continuity. The effect of availability issues can
be reduced by the use of multiple cloud provider services however creating
CHAPTER 5. DISCUSSION
81
and maintaining multiple application stacks in different clouds will be cumbersome. Distributed Denial of Services (DDoS) attacks can also add up to
the availability of services problem where the bot attacks can be established
on Cloud based SaaS implementations. Utility computing can absorb such
attacks and may also have core competency to protect systems for DDoS.
Data security and access: Most of the components of BSS can be deployed on cloud but data security and data access are the key concerns [61].
Public clouds are offered on public networks and so they are subject to the
attacks similar to any system on public network. The issue of auditability
[37] remains as well where necessary data access and regulations may impose
difficulties [44] while moving customer sensitive data to the cloud. As mentioned in thesis results, there was a recent case of Sony PlayStation where
customer sensitive data were stolen from the cloud. In such cases, operators
may lose their reputation and business. However, there are technologies like
virtual local area network (VLAN), encrypted storage and network middleboxes which can be used to protect data in public cloud as well as at the
time of data transfers to the public cloud [5]. In some cases, data encryption
before placing it on cloud may be even safer than the unencrypted data residing in an in-house data center. Hybrid cloud is also one way to ensure the
data remains in private cloud close to the other customer facing systems and
public cloud can be used only to process data with highly encrypted data
channels. Such architecture ensures data access and security very close to
the same of in-house local data storage.
Data transfer issues: Telecom applications, especially on BSS side, are
very data-sensitive and the data transfer rate as well as mechanism matters
a lot while integrating applications with other parts of the telecom systems.
Cloud providers charge 100$ to 150$ per terabyte of data transferred to
and from cloud. Also, there could be a possibility where large amount of
data transfer may take days if the network bandwidth is not high enough.
For example, to transfer 10 terabyte of data to the public cloud situated at
CHAPTER 5. DISCUSSION
82
another location in the same region takes 40-45 days with current bandwidth
capacity. Some simple solutions like sending 10 TB data disk via courier
services may save time and network cost [5], however telecom operators might
not want to take such risk of sending customer sensitive data over physical
transfer (i.e. courier or postal) services. Also, there are problems of low
performance with intra-networking technology for cloud where hypervisors
sometimes are not efficient enough to handle huge amount data transfer [25].
Legal requirements: There are other challenges too while moving MVNO
systems to the cloud in which legal requirements is one of the main challenges.
There are certain requirements set by government authorities which must be
followed when operating into telecom services domain. Standards for telecom infrastructure for data centers mention the basic SLA of 99.671% to
the fault tolerance SLA of 99.995% [38]. Such SLAs are difficult to manage on cloud as the cloud is not ready for such carrier-grade services. Also,
many countries have laws related to the customer sensitive data that it could
not be taken outside the national boundaries [43]. Some business may not
like the ability of other countries to control their customer data. However,
region-based cloud facilities make the situation less problematic in such cases.
For example, Amazon EC2 cloud instances are offered from US, Europe and
Asia-pacific regions. Legislation changes may require fast-paced customizations to the Telco applications along with changes in integration points etc.
and cloud makes it difficult to do such changes in a timely manner due to
application lock-in issues.
Performance bottlenecks: Telecom operators are not yet ready for
public cloud and at present they prefer only the private cloud as they believe
it will be difficult to manage performance and back to back SLAs with public
cloud. Real time monitoring and testing is not feasible if systems are implemented on cloud. In cloud computing environment, CPU and main memory
are shared well among virtualized instances but there are problems of sharing
I/O operations. There are problems with latency as well [41]. The experiment of measuring VOIP latency shows that in normal stand-alone systems
the latency is 150 milliseconds while on Amazon EC2 instance, the latency
grows up to 1000 milliseconds [38]. BSS applications rely on intensive I/O
operations and cloud may not be able to cope up with high risk throughput
[25]. Also there was an experiment conducted on EC2 instances showing
75% drop on available throughput and latency for intensive I/O operations
[7]. Extensive research is being carried out in area to reduce latency and
throughput issues on cloud; however major breakthrough is yet to come.
Integration and customization challenges: In BSS, other challenges
are the customization of BSS functionalities and integrations to the other
external systems. MVNOs and Telco try to down-size the integrations and
CHAPTER 5. DISCUSSION
83
integration points [36] and if Telco systems are moved to the cloud then they
will get more tightly-coupled systems. Most of the time MVNOs or Telco
update their services and functionalities and it takes 60-70% changes in a
specific systems sometimes to add or enhance functionality. For MVNOs its
very important to act fast and have the lowest time to market while delivering
services as quickly as possible. Cloud implementation makes it difficult due
to tight coupling and increase in integration points. So whether or not the
systems are on cloud, it must be easy and agile to make enhancements to the
system. Customer and data lock-in issues [1] also prevail on cloud computing
where it may be attractive to cloud provider to have such lock-ins and for
customers, it is not easy to move their data and applications effortlessly from
one cloud provider to another. The standardization of APIs is somehow
lacking in cloud environments.
Real-time operations: MVNO would be ready to have everything on
cloud but OSS layer is more difficult to implement on cloud for e.g. provisioning. Fault and performance management might not be implemented
on cloud which makes real time charging implementation on cloud difficult.
Also, fraud management systems are very critical and such systems need lot
of data from OSS as well as from BSS. Self-provisioning might be implemented on cloud but then service assurance would be difficult to manage.
5.4
Future Prospects
Cloud technology is evolving and different types of solutions are being introduced as per the industry specific needs. Telco clouds will provide carriergrade services within a particular region by offering highest level of data
security and performance. Hybrid clouds for Telco are viable and practical
solutions which will not require many modifications in current cloud architecture. Such concept will become norm for large enterprises where such enterprises will own private cloud to support their internal resources and they will
use public cloud in case of elastic needs [64] of computing resources. Community cloud architecture is also becoming famous where several organizations
and enterprises holding same industry or community build industry specific
clouds for themselves. In this approach, data security can be maintained by
not transferring data on public network and resource sharing will be costbeneficial. Hierarchical cloud architecture is already in place at some level
where Amazon provides EC2 instances at region or continent level. However, there will be a need to have such public or hybrid cloud infrastructure
at country or region level due to design and legal requirements especially in
telecom sector.
CHAPTER 5. DISCUSSION
84
CHAPTER 5. DISCUSSION
85
Chapter 6
Conclusion
MVNO systems mainly consist of BSS and marketing-sales systems. MVNO
buy the network capacity from MNO and use their OSS as well as networking
systems. Nowadays, there is a major shift in MVNO systems where MVNOs
tend to outsource most or all of their systems to focus more on their core
business competencies such as branding and reselling telecom services. For
MVNO business, cost-benefits and service level agreements with the partner companies are very important. The MVNO business model is generally
to drive price and service differentiation however there are certain MVNOs
driving geographical and market differentiations as well.
Cloud computing hype is providing the computer industry a new way of
doing things with old fundamentals. Several big companies like Amazon,
Google and Microsoft are providing cost efficient public hosting services as
utility computing service. Also, in-house private clouds enhance sharing of
resources and cost reductions efficiently. Hybrid cloud is created out of the
combination of public and private cloud computing architectures to efficiently
and cost-beneficially handle fluctuating loads [64] in the system. Cloud hosting models such as SaaS, PaaS and IaaS provide different delivery platform
to host and share applications and services on cloud.
In thesis, the main research question asked was related to application
mapping and evaluation of related parameters like performance and costbenefits while MVNO BSS on cloud. The action research was conducted as
part of the thesis in which Delphi method based expert interview was the
main method of research. There were two rounds of interviews conducted first
being off-line and second being online interviews. The results are analyzed
while ding system analysis, application mapping, cost-benefit calculations
and parameter evaluation.
Application mapping on the cloud analysis suggested that most of MVNO
systems shall be moved to the cloud with exception of prepaid and network
86
CHAPTER 6. CONCLUSION
87
(OSS) related systems. There were disagreements or doubts regarding implementation of Mediation platform on cloud as well. Along with the structured responses, the reasons behind the responses were also discussed and
mentioned in the thesis. Advantages like resource sharing, economy of scale,
SaaS based architecture and better performance will be driven by cloud in
context of MVNO BSS.
Cost-benefit analysis suggested that its possible to gain ROI improvements and reduction in IT admin cost while moving to the cloud. Also the
risk on investment will be minimized further however TCO will remain lower
due to the major outsourcing of MVNO systems on cloud. There will be
considerable transition from CAPEX to OPEX [15] along with significant
reduction in OPEX especially for MVNO BSS. Parameter evaluation suggested that cost-benefits and performance are important issues in MVNO
context while carrier-grade SLA may not be that important for MVNO as
they mostly deal with BSS. Performance, data security will be affected significantly while moving MVNO systems to the cloud while other parameters
such as SLA and customer service delivery may have minor impacts.
There were ideas discussed in the thesis related to MVNO systems and
cloud. One of the ideas was a proof of concept for implementing BSS like
enterprise systems on hybrid cloud and the other was providing billing-asa-service on cloud as PaaS. Billing-as-a-service will provide cost-effective,
ready-to-use and utility pricing based service to companies in requirement
of such solutions especially MNO and other startups. Implications were discussed for the thesis research in which data security, availability at some
extent, data transfer and legislation needs were the key concerns. Future
prospects were also discussed in the end in which ideas like application attribute mapping tools for application migration on cloud and Telco cloud
opportunities were the main points.
Bibliography
[1] Abu-Libdeh, H., Princehouse, L., and Weatherspoon, H.
RACS: a case for cloud storage diversity. In Proceedings of the 1st ACM
symposium on Cloud computing (New York, NY, USA, 2010), SoCC 10,
ACM, pp. 229240.
[2] Accenture, WSP and Microsoft. Cloud computing and sustainability: The environmental benefits of moving to the cloud. Tech. rep.,
Accenture, WSP and Microsoft, 2010.
[3] Ali, M. Green cloud on the horizon. In Cloud Computing, M. Jaatun,
G. Zhao, and C. Rong, Eds., vol. 5931 of Lecture Notes in Computer
Science. Springer Berlin / Heidelberg, 2009, pp. 451459. 10.1007/9783-642-10665-1 41.
[4] Amazon.
Amazon web services EC2 features,
2011.
http://aws.amazon.com/ec2/[Online; accessed 08-August-2011].
[5] Armbrust, M., Fox, A., Griffith, R., Joseph, A. D., Katz,
R. H., Konwinski, A., Lee, G., Patterson, D. A., Rabkin, A.,
Stoica, I., and Zaharia, M. Above the clouds: A Berkeley view of
cloud computing. Tech. Rep. UCB/EECS-2009-28, EECS Department,
University of California, Berkeley, Feb 2009.
[6] Badger, L., Grance, T., Patt-Corner, R., and Voas, J. Draft
cloud computing synopsis and recommendations. Tech. Rep. NIST Special Publication 800-146, National Institute of Standards and Technology, 2011.
[7] Barker, S. K., and Shenoy, P. Empirical evaluation of latencysensitive application performance in the cloud. In Proceedings of the first
annual ACM SIGMM conference on Multimedia systems (New York,
NY, USA, 2010), MMSys 10, ACM, pp. 3546.
88
BIBLIOGRAPHY
89
BIBLIOGRAPHY
90
[20] Dillon, T., Wu, C., and Chang, E. Cloud computing: Issues
and challenges. In Advanced Information Networking and Applications
(AINA), 2010 24th IEEE International Conference on (april 2010),
pp. 2733.
[21] Dittberner. OSS/BSS knowledgebase. Tech. rep., Dittberner, Nov.
2010.
s, I., and Skog, R.
[22] Gabrielsson, J., Hubertsson, O., nacio Ma
Cloud computing in telecommunications. Tech. Rep. 1, Ericsson, 2010.
[23] Grossman, R. The case for cloud computing. IT Professional 11, 2
(march-april 2009), 2327.
[24] Hajjat, M., Sun, X., Sung, Y.-W. E., Maltz, D., Rao, S.,
Sripanidkulchai, K., and Tawarmalani, M. Cloudward bound:
planning for beneficial migration of enterprise applications to the cloud.
In Proceedings of the ACM SIGCOMM 2010 conference on SIGCOMM
(New York, NY, USA, 2010), SIGCOMM 10, ACM, pp. 243254.
[25] Huang, W., Liu, J., Abali, B., and Panda, D. K. A case for
high performance computing with virtual machines. In Proceedings of
the 20th annual international conference on Supercomputing (New York,
NY, USA, 2006), ICS 06, ACM, pp. 125134.
[26] Joshipura, A. The cloud: what operators stand to gain or lose. Tech.
rep., Ericsson, 2010.
[27] Keber, B. Applying eTOM to public utilities industry. Tech. rep.,
Marand d.o.o., 2004.
[28] Kelly, M. B. Report: The telemanagement forums enhanced telecom
operations map (eTOM). Journal of Network and Systems Management
11 (2003), 109119. 10.1023/A:1022449209526.
[29] Kelly, M. B. Enhanced telecom operations map (eTOM)-the business
process framework. Tech. rep., Tele Management Forum, 2007.
mma
inen, H. Mobile virtual network operator
[30] Kiiski, A., and Ha
strategies: Case Finland. In Proceedings of the International Telecommunication Society (ITS) 15th Biennial Conference (Berlin, Germany,
Sept. 2004).
BIBLIOGRAPHY
91
[31] Kondo, D., Javadi, B., Malecot, P., Cappello, F., and Anderson, D. P. Cost-benefit analysis of cloud computing versus desktop
grids. Parallel and Distributed Processing Symposium, International 0
(2009), 112.
[32] Koutsopoulou, M., Kaloxylos, A., Alonistioti, A., Merakos,
L., and Kawamura, K. Charging, accounting and billing management schemes in mobile telecommunication networks and the internet.
Communications Surveys Tutorials, IEEE 6, 1 (quarter 2004), 5058.
[33] Liebowitz, S. J., and Margolis, S. E. Network externality: An
uncommon tragedy. The Journal of Economic Perspectives 8, 2 (1994),
pp. 133150.
[34] Luoma, E., Frank, L., and Pulkkinen, M. Overview of telecom operator software market. In Vertical Software Industry Evolution,
W. A. M
uller, M. Bihn, P. Tyrvainen, and O. Mazhelis, Eds., Contributions to Management Science. Physica-Verlag HD, 2009, pp. 3542.
10.1007/978-3-7908-2352-3 5.
[35] Maes, S. Understanding the relationship between SDP and the cloud.
CLOUD COMPUTING 2010, The First International Conference on
Cloud Computing, GRIDs, and Virtualization (Nov. 2010), 159163.
[36] Mazhelis, O., Tyrvainen, P., and Viitala, E. Modeling software integration scenarios for telecommunications operations software
vendors. In Industrial Engineering and Engineering Management, 2007
IEEE International Conference on (dec. 2007), pp. 4954.
[37] Mell, P., and Grance, T. Effectively and securely using the cloud
computing paradigm. Tech. rep., NIST, July 2010.
[38] Murphy, M. Telco cloud. Cloud Asia Conference (Presentation) (May
2010).
[39] Networks, N. S. Cloud computing business boost for communications industry. Tech. rep., Nokia Siemens Networks, 2011.
[40] Nokia-TietoEnator. OSS/BSS reference architecture and its implementation scenario for fulfillment. Tech. rep., Nokia-TietoEnator, 2004.
[41] Paivarinta, R., and Raivio, Y. Performance evaluation of NoSQL
cloud database in a telecom environment. Closer (May 2011).
BIBLIOGRAPHY
92
BIBLIOGRAPHY
93
[53] Sripanidkulchai, K., Sahu, S., Ruan, Y., Shaikh, A., and Dorai, C. Are clouds ready for large distributed applications? SIGOPS
Oper. Syst. Rev. 44 (April 2010), 1823.
[54] SugarCRM. Sugarcrm sizing guidelines. Tech. rep., SugarCRM Inc,
2010.
[55] Ullman, J., Osborne, I., Taylor, M., and Endre, R. TCS innovation forum: Cloud computing. Tech. rep., Tata Consultancy Services,
2011.
[56] Ulset, S. Mobile virtual network operators: a strategic transaction
cost analysis of preliminary experiences. Telecommunications Policy 26,
9-10 (2002), 537549.
[57] Varoutas, D., Katsianis, D., Sphicopoulos, T., Stordahl, K.,
and Welling, I. On the economics of 3G mobile virtual network
operators (MVNOs). Wireless Personal Communications 36 (2006),
129142. 10.1007/s11277-006-0027-5.
[58] Von den Hoff, K., Opitz, M., et al. Cost reduction in the telecom
industry. Tech. rep., Arthur D. Little, 2010.
[59] Wyld, D. C. The utility of cloud computing as a new pricing and
consumption - model for information technology. International Journal
of Database Management Systems (IJDMS) 1, 1 (Nov. 2009).
[60] Wyld, D. C. The cloudy future of government IT: Cloud computing
and the public sector around the world. International Journal of Web
& Semantic Technology 1, 1 (2010), 120.
[61] Wyld, D. C. Risk in the clouds?: Security issues facing government use
of cloud computing. In Innovations in Computing Sciences and Software
Engineering, T. Sobh and K. Elleithy, Eds. Springer Netherlands, 2010,
pp. 712. 10.1007/978-90-481-9112-3 2.
[62] Yin, R. K. Case Study Research : Design and Methods, 3rd ed. Applied
Social Research Methods. SAGE Publications, Dec. 2002.
[63] Yoo, C. Cloud computing: Architectural and policy implications. Review of Industrial Organization 38 (2011), 405421. 10.1007/s11151011-9295-7.
BIBLIOGRAPHY
94
[64] Zhang, H., Jiang, G., Yoshihira, K., Chen, H., and Saxena,
A. Intelligent workload factoring for a hybrid cloud computing model.
In Services - I, 2009 World Conference on (july 2009), pp. 701708.
[65] Zhang, Q., Cheng, L., and Boutaba, R. Cloud computing: stateof-the-art and research challenges. Journal of Internet Services and Applications 1 (2010), 718. 10.1007/s13174-010-0007-6.
[66] Zoutendijk, R. Strategic overview of cloud computing adoption. Tech.
rep., IBM Netherland B.V., 2010.
Appendix A
Appendix
A.1
Instructions:
Interviewee should have clear picture of the background, purpose and
work to be conducted as part of thesis research. Interviewee should
have received research proposal beforehand.
Also interviewee should know about the research method, interview
formation and research design in brief for the thesis research.
All the details of interviewee and business critical information imparted
from interviewee should remain anonymous throughout the process and
thesis research.
Notes will be taken during interview. Also audio or video will be taken
as per the choice and convenience of interviewee.
Questions:
1. How does MVNO system work? (The main focus is on Business Support Systems). What are the integral parts of the system? Please
explain the architecture in brief. Also explain MVNO business and
operations in brief.
2. How are high performance, carrier-grade SLA, cross-location architecture (if applicable) and cost-benefits important in current MVNO operation context?
95
APPENDIX A. APPENDIX
96
Integration
Processing
Bills
Collecting
Payments
MVNO Systems
Self-service
(end-user facing
web-portal)
CRM (Customer
Relationship
Manager)
OMS
(Order
Management
System)
Billing & CRM
Integrator (Engine or JMS implementation)
Product Catalog
(Repository of
products)
Mediation (processing
call
records)
Rating & Invoicing (rating calls
and preparing
bills)
Bill Formatter
(Formatting and
printing bills)
Account Receivable (receiving
payments)
Deployment
on cloud
Reason for
deployment
Comments
APPENDIX A. APPENDIX
Inventory
Prepaid Operation
97
Collection (due
payments, barring, processing
payments)
Resource Management (managing
SIM,
MSISDN, Data
number
resources)
Prepaid
Card
System
(real
time charging)
Table A.1: Mapping of MVNO systems on cloud
4. How and if following parameters are affected while migrating some part
or the whole of the MVNO system to the cloud?
Parameter
Yes/No/NA
How?
Performance (Throughput, Latency)
Service Level Agreement (SLA)
Service Delivery
Customer Services
Security (N/w and Data Security)
Carbon Footprints
Business Model changes
Cross-location Architecture
Table A.2: Parameters affected while moving MVNO systems to cloud
5. How much resources are needed at present to run the system on monthly
basis? (Processing power, memory, network bandwidth etc). Will it be
feasible to move to the cloud?
APPENDIX A. APPENDIX
Customer
base (No of
Customers)
ARPU
(Average
Revenue
per User)
98
Processing
(in monthly
hours)
DB Storage
(in MB)
NW Bandwidth
(in
Gbps)
6. Which cost factors are involved in telecom systems especially for MVNOs?
How do you calculate pricing for telecom cloud? How do you get Return
on Investment (ROI) while moving or implementing MVNO systems on
cloud?
7. How do you improve the cost-benefits for MVNO system? How are the
high-level calculations made to derive cost-benefits? Cloud may drive
the cost-benefits in MVNO context. Are you willing to accept cloud
only on the basis of cost-benefits?
8. What are the implications of cloud deployment of MVNO systems?
How government laws can affect the deployment? What are the financial implications?
9. What do you think of the future of MVNO and Telecom Software industry also considering importance of cloud? What vision do you look
for?
10. Do you think the idea Billing as a Service is viable and feasible? Do
you think all systems at one place rather than in different places is a
good idea? (Interviewer will explain Billing as a Service using cloud
for MVNO)
Industry specific questions:
11. How is the company coping up with the idea of Cloud Computing and
especially Telco Cloud in their operations? How are companys R&D
operations aligned to manage research related to the cloud?
12. Does the company have strategy to move/migrate their systems or
part of the system to the cloud? Does the company see long-term
cost-benefits gained while moving Telco Systems to the cloud?