273

13

Techno-economic and Business Feasibility Analysis of

5G Transport Networks
Forough Yaghoubi 1 , Mozhgan Mahloo 2 , Lena Wosinska 1,3 , Paolo Monti 1,3 ,
Fabricio S. Farias 4 , Joao C. W. A. Costa 4 , and Jiajia Chen* ,1,3
1
School of Electrical Engineering and Computer Science, KTH Royal Institute of Technology, Kista, Sweden
2
ÅF Technology, Digital Solutions Division, Stockholm, Sweden
3
Department of Electrical Engineering, Chalmers University of Technology, Gothenburg, Sweden
4
School of Electrical and Computer Engineering, Institute of Technology of the Federal University of Pará, Belém, Brazil

13.1 Introduction
The exponential growth of mobile data traffic, mainly driven by multimedia services and
an increase in the number of connected devices, brings new challenges for mobile net-
work operators (MNOs) [see Cisco - 2015; UMTS Forum - 2011; Osseiran et al. - 2013;
Zander and Mähönen - 2013]. Traditionally this capacity growth has been addressed by
looking for new spectrum opportunities, enhancing spectrum efficiency, and/or adding
macro cell sites. However, the spectrum is not an infinite resource, and its efficiency is
improving at a much slower pace than the increase in the capacity demand [see Femto
forum - 2010]. It is also complicated to acquire new base station (BS) sites in urban areas,
not to mention the inefficiency of using macro sites to serve users that mainly reside
indoors [see Norman - 2010; Ericsson AB - 2013]. A promising way to solve this capac-
ity crunch is to deploy heterogeneous networks (HetNets), where high-power macro
cells provide coverage, and less expensive outdoor/indoor small cells are deployed close
to the end user to ensure that capacity is provided only where it is needed.
The benefits of HetNets over homogeneous deployments (i.e. macro cells only) have
been demonstrated in a number of studies in terms of both cost and power consumption
[see Markendahl et al. - 2008; Markendahl and Mäkitalo - 2010; Aleksic et al. - 2013;
Khirallah and Rashvand - 2011; Claussen et al. - 2008]. The authors of [Markendahl
et al. - 2008] show that using only small cells, especially for indoor coverage, can decrease
the deployment cost of radio access networks (RANs) by nearly a factor of five. The study
in [Markendahl and Mäkitalo - 2010] emphasized that small cell deployments are a cost
efficient alternative to macro densification, especially in scenarios where the capacity
demand is high (e.g., beyond 100 Mbps km−2 ). The work in [Aleksic et al. - 2013] evalu-
ates the benefits of indoor small cells on improving the capacity available for residential
and enterprise users, while [Khirallah and Rashvand - 2011] discuss the energy reduction

* Corresponding Author: Jiajia Chen; jiajiac@kth.se

Optical and Wireless Convergence for 5G Networks, First Edition.

Edited by Abdelgader M. Abdalla, Jonathan Rodriguez, Issa Elfergani, and Antonio Teixeira.
© 2020 John Wiley & Sons Ltd. Published 2020 by John Wiley & Sons Ltd.
274 13 Techno-economic and Business Feasibility Analysis of 5G Transport Networks

that mobile operators can achieve thanks to HetNet deployments. Finally, the work in
[Claussen et al. - 2008] shows that with HetNet deployments it is possible to achieve
energy saving up to 60% in urban areas. On the other hand, the introduction of small cells
has an impact on the backhaul network [see Farias et al. - 2013a; Tombaz et al. - 2011],
which is responsible for collecting data traffic at the BSs (i.e. after baseband processing)
and for sending it to the metro/aggregation segment. With HetNets the backhaul seg-
ment tends to become more complex compared to homogeneous wireless deployment
solutions [see Juniper Networks - 2011]. This is because a large number of links are
required to aggregate the data traffic generated by all the small cells, each one working
at a peak rate at least tens of Mbps. These rates can hardly be guaranteed with legacy cop-
per based infrastructures. in particular not over a long distance [see Farias et al. - 2013b].
Such a scenario is likely to force mobile operators to upgrade their backhaul networks
to avoid potential bottlenecks in terms of capacity.
In a competitive telecom market, where new services and technologies are frequently
appearing, and the revenue margin is constantly reducing [see Giles et al. - 2004], for
each new deployment/upgrade all the possible cost drivers need to be carefully consid-
ered. Knowing that the cost of the backhaul segment is already a not negligible part of the
total cost of ownership (TCO) in homogeneous wireless networks [see Geitner - 2005], it
can be expected that with an increasing number of small cells (i.e. as in a HetNet deploy-
ment) the impact of the backhaul segment on the TCO of RANs will become even more
critical [see Skyfiber - 2013]. Therefore, finding cost efficient backhaul solutions have
become an important challenge in recent years. The rationale is simple: if not correctly
predicted, the extra backhaul cost might reduce the benefits brought by the deployment
of small cells.
On the other hand, restricting the analysis of the impact of backhaul only to TCO
assessment is risky. TCO can be used to compare the various backhaul alternatives, but
it provides estimates only about the cost aspects of a given technology/architecture.
This means that from a TCO analysis it is not possible to extrapolate any information
about the profitability of a specific backhaul solution, (i.e. the amount of money gained
with respect to the investment). These considerations include many other factors such
as yearly revenues, user penetration rates, the number of competitors in the area, and
regulations, just to name a few. As a result, profitability can only be assessed with a
comprehensive economic feasibility framework that in addition to TCO is also able to
provide estimates about the net present value (NPV) of a backhaul deployment [see
Cid et al. - 2010].
There are some attempts in the literature to provide cost modeling of mobile net-
works. [Frias and Pérez - 2012] numerically studied the cost savings achievable with the
deployment of indoor small cells in regards to RAN deployments. The work in [Ahmed
et al. - 2013] studies various wireless architectures (both homogeneous and heteroge-
neous) and tries to assess the impact of backhaul on the entire TCO. An author in [Frias
and Pérez - 2012] proposed a comprehensive methodology to analyze the TCO of a
number of backhaul network options based on fiber, copper, and microwave. The use
case under examination considers a European urban scenario with both outdoor and
indoor users. The latter is served by a layer of femto cells deployed inside the buildings
where they reside, while the former are catered by macro BSs. However, many details
are ignored that might affect the results in terms of cost saving. The authors in [Soh
et al. - 2003; Kuo et al. - 2010] compared different microwave based backhaul topologies
 13.2 Mobile Backhaul Technologies 275

(including mesh and tree) with respect to their total cost. They reached the conclusion
that mesh structures are a cost efficient option for homogeneous wireless deployment,
but they did not consider HetNet deployments.
The study presented in [Senza fili - 2011] compares the deployment cost of back-
hauling in a long term evolution (LTE) homogeneous wireless network using fiber
and microwave technologies. This study concludes that for scenarios with low cell
density, the microwave is the cheapest option. However, only a homogeneous wireless
deployment was evaluated in this work. The works in [Monti et al. - 2012; Tombaz
et al. - 2014] compare different technology and topology options for backhauling a
HetNet deployment, but their main focus is only on energy consumption. The work
in [Mahloo et al. - 2014b] introduces for the first time a general TCO evaluation
methodology of mobile backhaul networks, including a detailed breakdown of the
capital expenditure (CAPEX) and operational expenditure (OPEX). On the other
hand, no NPV analysis is provided. From the list of research efforts described so far,
it becomes clear that there is a need for an assessment framework that in addition to
mere cost evaluation also estimates the NPV of a given backhaul deployment. This issue
has been addressed in [Yaghoubi et al. - 2018] where the authors provided complete
economic feasibility evaluation considering TCO and NPV calculation.
This chapter introduces a techno-economic framework that provides a complete mar-
ket analysis of the various business actors for any type of mobile access network deploy-
ments (including both the homogeneous and the heterogeneous cases). The aim of the
proposed model is to advise operators on what type of backhaul investment should be
made, at which point in time, and based on what technology, in order to maximize their
profits. This is done by providing a complete cash flow and NPV evaluation on top of a
detailed TCO estimation. It should be noted that in techno-economics, the NPV is the
most important criteria for understanding whether a network deployment is profitable
or not. Finally, a sensitivity analysis of several important cost factors is also included,
showing the influence of uncertainty on the assumptions and input values used in the
case studies.

13.2 Mobile Backhaul Technologies

The backhaul network is responsible for aggregating the users’ traffic from the wire-
less access (i.e. BSs) to the metro/backbone segment of the network. Operators may
select from more than one technology for their backhaul network(s) according to their
needs in terms of capacity, reliability, cost, and expected deployment duration. In gen-
eral, backhaul networks may be based on copper, fiber, microwave, or a combination of
all these technologies (i.e. hybrid backhaul) [see Ercisson AB - 2014]. Recent backhaul
technologies such as millimeter wave [see Nie et al. - 2013] and free space optic [see Feng
et al. - 2014] have been evaluated in the literature and were introduced to the market,
but they are not yet mature enough for deployment on a massive scale.
Currently, microwave represents nearly 50% of all backhaul deployments, and it is
expected that it will maintain this share of the market in the years to come [see Ercisson
AB - 2014]. This is mainly due to the moderate installation cost and its relatively short
time to deploy. With the most recent technological development, it is possible to have
microwave links operating at 1 Gbps up to few kilometers [see Ercisson AB - 2014;
276 13 Techno-economic and Business Feasibility Analysis of 5G Transport Networks

Figure 13.1 Microwave based backhaul

architecture.

Macro BS

MW antenna

MW hub

Metro Network

Coldrey et al. - 2013]. Regardless of the topology in which it is organized (i.e. mesh, tree,
ring, star, or any combination of them), a microwave based backhaul consists of several
point-to-point (P2P) or point-to-multipoint (P2MP) links, each one requiring antennas
at every endpoint. In the case of a P2P link, which is the major focus of this chapter, one
antenna is located at the BS (or on top of a building), while the other one is connected
either to a switch at the first aggregation point of the backhaul infrastructure (in the
case of a multi-stage backhaul) or directly to a switch at the metro/aggregation node.
Where several microwave antennas are co-located in one place, a tower mast (i.e. also
referred to microwave hub) needs to be installed. When possible microwave antennas
can be also installed on the same tower as the macro base station antennas. Figure 13.1
shows an example of a simple microwave based backhaul architecture.
Copper-based backhaul segments amount approximately to 20% of all current exist-
ing backhaul deployments [see Tipmongkolsilp et al. - 2011; Ercisson AB - 2017]. Most
likely, they will be gradually replaced by other technologies due to their limited ability to
provide high capacities over long distances (i.e. more than 100 Mbps can be guaranteed
only up to 300 m [see Farias et al. - 2013b]). However, there is some new advancement
in copper technology that may lead to its survival for several more years before it is
replaced totally by other technologies. One example is the G.fast standard that is cur-
rently developed by ITU-T and aims at achieving bit rates up to 1 Gbps but still over
short distances [see Lins et al. - 2013; Nokia - 2018].
Fiber based backhaul can provide ultra-high capacity over long distances. However, it
is relatively time-consuming and expensive to deploy a fiber infrastructure, especially if
it is done from scratch (i.e. greenfield deployment). On the other hand, in places where a
cable infrastructure (e.g., ducts from a previous copper installation) is already available,
faster deployment of fiber is possible [see Farias et al. - 2016].
Fiber access networks can be deployed as P2MP topology or as P2P interconnection.
In the latter case, one optical line terminal (OLT), a device corresponding to the service
provider endpoint, located in the central office (CO) is connected to an optical net-
work unit (ONU), the endpoint at the user/wireless network side, via a dedicated fiber
link (Figure 13.2a). In a P2MP architecture (Figure 13.2b), each OLT is connected to
several ONUs via a passive splitter (i.e. in the case of passive optical networks (PONs)),
 13.2 Mobile Backhaul Technologies 277

ONU
CO
Fibers ONU
(a)
OLT ONU

ONU
CO cabinet
Fiber
Fiber Switch/ Fiber ONU
(b)
splitter Fiber
OLT
ONU

Figure 13.2 Fiber based backhaul architecture.

or through an ethernet switch (i.e. in the case of active optical networks). Splitters and
ethernet switches are located in a node that is referred to as the remote node. Thanks
to the ability of fiber to deliver high capacity over long distances (i.e. in the order of sev-
eral tens of kilometers), a backhaul network based on the optical transmission makes it
possible to connect cells directly to the mobile core network, without any intermediate
stages. This is not the case for a microwave based backhaul network where the con-
nectivity between the microwave hub and the mobile core network typically requires a
fiber link (Figure 13.1). Fiber cables are often installed inside ducts that are buried under
the ground. This process is referred to as digging (trenching) and it represents the most
expensive part of a fiber network deployment.
To achieve better utilization of the radio resources some operators prefer to use a cen-
tralized radio access network (C-RAN) architecture [see Chih-Lin et al. - 2014]. Unlike
traditional RANs, where the radio unit and the baseband processing unit (BBU) are
co-located at the base station site, in a C-RAN architecture, BBUs are decoupled from
the BSs and centralized into one or more BBU hotel(s). In a C-RAN architecture, the
transport network is divided into two parts: fronthaul and backhaul. The fronthaul seg-
ment is responsible for transporting the traffic between the remote radio units (RRUs)
and the BBU hotels, while the backhaul part provides transport connectivity between
the BBU hotels and the mobile core network. The communication between RRUs and
BBUs, which can be based on, e.g., the common public radio interface (CPRI) protocol
[see CPRI - 2013], normally requires capacities in the order of tens of Gbps, making opti-
cal transmissions (i.e. in particular, radio over fiber techniques) the perfect candidate for
fronthaul.
In summary, there is a consensus that fiber and microwave are the two main candi-
dates for backhauling current and future mobile network deployments. For this reason,
the case study carried out in this chapter focuses on these two technologies. In addi-
tion, the framework proposed in this chapter addresses specifically only the backhaul
segment (i.e. no C-RAN), but it is general enough to also be applied to the other 5G
transport solutions (e.g., fronthaul).
278 13 Techno-economic and Business Feasibility Analysis of 5G Transport Networks

13.3 Techno-economic Framework

This section first discusses the role played by business viability assessment in network
deployments, and then it presents a framework that can be used by mobile network
operators to estimate the profitability of a given backhaul deployment.
Figure 13.3 presents the life cycle of a communication network, which typically con-
sists of four phases: planning, initial installation, operational phase (e.g., providing con-
nections to customers, keeping a network up and running), and teardown. The planning
phase takes place before any new deployment. This is the most crucial step for under-
standing if a deployment is feasible or not, and to reduce the risk of the investment
not bringing profit. The reason is the following. Even if a technology is already mature
enough to be deployed, the market may not be ready for it, e.g., the user penetration
might be too low, or potential users may not be willing to pay extra for a specific service.
All these aspects need to be assessed via a comprehensive techno-economic and risk
analyses framework in order to validate the economic viability of a new deployment.
More specifically, it is crucial to quantify the total expenses required during the network
lifetime, as well as the estimated revenues and cash flows.
This information should then be used to estimate the payback period, i.e. the time
required for the return of the investment. If the payback period is too long, or the total
cash flow is negative, it is not advisable (from a purely economic point of view) to carry
out the project.
If the results from the planning phase show that a given deployment is profitable, then
the initial installation phase starts. During this phase, operators sustain an upfront cost,
money that is typically considered as part of the CAPEX in the TCO calculations. Once
deployed, the network needs to be kept up and running (i.e. the operational phase). All
the expenses that occur in this phase are considered part of the OPEX.
Finally, when the current network needs to be replaced/upgraded (e.g., new technol-
ogy is ready to be deployed), the customers will gradually be subscribed to the new net-
work. This period is referred to as the teardown phase, and it ends when the migration
to the new service(s) is completed. The expenses related to the network teardown phase
include mainly labor cost, and they typically represent a relatively small part of the TCO
compared to the money spent during the installation and the operational phase. For
this reason, the impact of the teardown phase can be, in a first approximation, ignored
in most cases.
As it was already explained, the profitability of a given deployment is a crucial point
in deciding whether or not a new deployment should take place. For this, a complete
Figure 13.3 Network life cycle.
Planning

Initial
Teardown
installation

Operational
phase
 13.3 Techno-economic Framework 279

Figure 13.4 Techno-economic

framework.
Architecture
module

Market
Topology
module
module

Network
dimensioning
tool

Business
Cost module
module

Total cost of
owner ship
(TCO) module

Techno-economic results

techno-economic framework analysis is required. Figure 13.4 presents an assessment

methodology (consisting of a techno-economic market and risk analysis) that can be
used by mobile operators to analyze the business feasibility of a given backhaul deploy-
ment. This framework consists of several modules, each one presented in the following
sections where the required input and the expected output will be described in detail.

13.3.1 Architecture Module

The objective of this module is to define the technology used in the backhaul segment
together with the type of components to be installed in each location. For example, in
the case of a microwave-based backhaul, antennas are required in both sides of the
microwave links, while in the case of fiber-based backhaul components such as OLT,
splitting devices, and ONUs need to be installed at central offices, remote nodes, and
user premises, respectively.

13.3.2 Topology Module

The network topology defines the way in which the various components of a given
architecture are interconnected. Examples of network topologies are ring, star, tree,
and mesh. Another important parameter included in the topology module is the demo-
graphical data of the region that is under study. The number of buildings, user density,
size of the geographical region, existing infrastructure (e.g., available ducts) are also an
input for the topology module. In terms of output, the topology module can compute
the number of network nodes (e.g., COs, remote nodes, cabinets), their locations, the
distances between different nodes, and the equipment type that should be installed in
each location. These parameters are then provided to the network dimensioning tool.
280 13 Techno-economic and Business Feasibility Analysis of 5G Transport Networks

13.3.3 Market Module

When planning a network it is crucial to consider market related data such as user pen-
etration rate, operator’s market share, user behavior, service prices, user churn rate (i.e.
percentage of user expected to unsubscribe from a service), area throughput, quality of
service (QoS), and connection availability, just to name a few. These parameters are fed
to the market module, which in turn can estimate the possible revenues, the number of
users that are expected to subscribe to and unsubscribe from the service under exam.
The outputs of the market module are then provided to the dimensioning tool.

13.3.4 Network Dimensioning Tool

By processing the inputs received from the architecture, topology, and market mod-
ules the dimensioning tool calculates, for a given scenario, the amount of required new
infrastructure (e.g., fibers, ducts, hubs) and the number of components needed in the
various network locations on a yearly basis. Moreover, the dimensioning tool also cal-
culates the value of some operational parameters related to the labor activities (e.g.,
traveling time to a certain network node for reparation).

13.3.5 Cost Module

A cost module is important for understanding how the various cost parameters in a
TCO module vary with time. For example the price of a specific component normally
decreases as a function of the (increasing) production volumes, the quantity purchased
on the market, and the level of maturity of the technology. On the other hand, the
expenses related to human resources (e.g., technician salaries) typically increase each
year. Therefore, price variation should be considered while calculating the network
expenses. Price erosion in time can be calculated via a learning curve that is used in the
industry to predict the reduction of the cost of a product [see Verbrugge et al. - 2009].
However, finding the right learning curve is not an easy task. The cost module applied
in the proposed framework refers to a widely used linear formula for calculating the
cost variation, presented below:
Pj = P0 + 𝛼Pj−1 (13.1)
where Pj denotes the price in a year j of network lifetime, and P0 is the price at the
beginning of the project. The coefficient 𝛼 denotes the cost change factor. This parameter
has a negative value when calculating how the price for hardware components varies (i.e.
this price normally decrease with time). On the other hand, 𝛼 has a positive value when
calculating how the price varies in time for parameters not related to the hardware, e.g.,
salaries and energy costs. In reality 𝛼 might also vary with time, however for simplicity,
in this study, 𝛼 is considered to be constant during the whole network lifetime.

13.3.6 Total Cost of Ownership (TCO) Module

This section presents the TCO module used in the proposed framework. The module
covers both the CAPEX and the OPEX aspects of the backhaul segment. More specifi-
cally, the module includes all the costs incurred during the backhaul lifetime (i.e. from
 13.3 Techno-economic Framework 281

Total cost of owner ship (TCO)

CAPEX OPEX

Fault Spectrum & fiber

Infrastructure Equipment management leasing

Purchasing Installation Energy cost Floor space

Maintenance

Figure 13.5 Cost classification of a TCO module.

the network deployment phase, when an upfront investment is required, up to all cost
aspects related to the operational processes).
Figure 13.5 presents the cost classification according to the proposed cost module.
Since, in general, a backhaul segment may comprise more than one technology (i.e. a
hybrid architecture), the proposed module accounts for the presence of both fiber and
microwave. The details of each part are presented next.

13.3.6.1 Capital Expenditure (CAPEX)

The CAPEX refers to all the expenses related to the backhaul network deployment cost.
According to the model in Figure 13.5, CAPEX can be divided into two main parts, i.e.
equipment and infrastructure costs. They are described in the following.

Equipment cost The equipment cost is the sum of all expenses related to purchasing
the backhaul components, i.e. according to the results of the dimensioning tool, and
to install them in their specific locations.

Infrastructure cost The total infrastructure cost of a mobile backhaul segment corre-
sponds to the investment needed to deploy the fiber infrastructure as well as the cost
of leasing fibers (when the fiber infrastructure has already been deployed by other
providers and is available for leasing). It also includes the expenses needed to install the
microwave hubs, i.e. masts and antennas, where needed. The fiber infrastructure cost
includes all the expenses related to trenching, purchasing of fiber cables, and pumping
fibers into the ducts. Trenching can be defined as placing the optical fibers inside the
ducts that are buried under the ground. In many cases, MNOs prefer to lease fibers
instead of deploying their infrastructure. In such a case, the infrastructure cost includes
an upfront charge per kilometer of leased fiber paid to the infrastructure owner.

13.3.6.2 Operational Expenditure (OPEX)

OPEX refers to the expenses occurred during network operations over a predefined time
interval (i.e. the network operational time). The main OPEX components are indicated
in Figure 13.5 and they are defined below.
282 13 Techno-economic and Business Feasibility Analysis of 5G Transport Networks

Spectrum and fiber leasing This cost refers to the fee that should be paid to lease
microwave spectrum or fiber infrastructure. When leasing fibers, an MNO is charged
a yearly fee for the maintenance and reparation of the rented fibers in addition to the
upfront expenses. The yearly cost of spectrum leasing for a licensed microwave link
varies depending on channel capacity (i.e. class i) and the frequency band.

Energy cost The electricity bill is part of the OPEX. This cost is obtained by summing
up the energy cost of all the active equipment in the various backhaul locations (i.e. CO,
cabinets, microwave sites, and equipment placed inside buildings).

Maintenance cost A regular maintenance routine is needed to keep a backhaul network

up and running. This includes monitoring and testing the equipment, updating soft-
ware (including renewing licenses when needed), and replacing deteriorated compo-
nents (e.g., batteries). The total maintenance cost reflects the maintenance cost of central
offices, cabinets, and microwave links. In order to ensure that the network and all the
services are running as expected, full-time monitoring is also required. This translates
into extra monitoring expenses in the total maintenance cost of the network. Operators
consider several rounds of maintenance procedures for each central office depending
on the number of users and services covered by each one of them. Microwave links also
require regular monitoring, because antennas might tilt and lose their line of sight.

Fault management Fault management refers to the expenses related to the reparation of
failures that might occur in a backhaul network. The total yearly reparation cost for the
backhaul network is defined as the sum of the reparation cost of each failure occurring
during the year as well as the penalty paid to the users as specified in the service level
agreement (SLA).
The reparation cost depends on the cost of replacing a failed component (if needed)
in each year, the mean reparation time of each device, and the time to travel to the
location of the failure. The penalty quantifies the fine that operators need to pay to
the customers when the service interruption is longer than the threshold defined in
the SLA, TSLA . Let t show a period where TSLA needs to be satisfied, which can be a
year, a month or even a day. If the mobile backhaul has a failure in this period t, one
or more macro cells might be out of service, and potentially a large number of cus-
tomers can lose connectivity. Therefore, we can consider penalty costs for a backhaul
provider when the macro cell backhaul connectivity is lost due to a failure. NjMac and
unAvij denote the number of macro cells with high importance in year j and the con-
co∕h
nection unavailability of the backhaul link to the macro cell i. The penalty rate Pi
is agreed in the SLA and is dependent on the importance of service outage for the
customer.
N Mac
∑
Ln
∑
j
co∕h
Penalty = Pi (unAvij − TSLA ). (13.2)
j=1 i=1

The above equation is valid if the service outage time is larger than TSLA , otherwise the
penalty cost is equal to zero.

Floor space cost The floor space cost is a yearly rental fee paid by an operator to house
its equipment, i.e. to place components in racks with standard size in various locations.
 13.3 Techno-economic Framework 283

13.3.7 Business Models and Scenarios

This module accounts for the business-related parameters, i.e. which are the actors
in the area, the co-operation models between actors and the various governmental
entities (e.g., municipalities). Business actors can be categorized into the following:
physical infrastructure provider, a network provider, a service provider, and the MNO.
Other important business-related parameters are the market share of each operator,
open access models (if any), and regulations with respect to sharing the infrastructure.
For example, if inside a commercial building it is not possible to install multiple
independent networks, two mobile operators may agree to share the infrastructure.
Another possibility is for the incumbent to offer roaming service to any newcomer
wanting to support its customers inside that building. Some possible business cases
related to backhaul deployments are mentioned below.
• Case 1: an operator provides both mobile and fixed network services to its customers.
In this case, the operator could reuse part of the fixed network infrastructure for back-
haul services. In this way, the cost of the backhaul infrastructure (which is a relevant
part of the backhaul TCO) can be reduced, and the period for the return on the invest-
ment will be shorter.
• Case 2: a mobile operator pays to a fixed network provider for backhaul connectivity
to each base station.
• Case 3: a mobile operator deploys its backhaul network independently.
In the case of backhaul for small cells deployed indoors, the two main business models
considered in the literature are the closed subscriber group and the open subscriber
group. In the former case only a closed group of users can access the indoor cells (i.e.
it is considered as private network for improving the quality of the service) while in the
latter case everyone in the range of the cell can connect [see Frias and Pérez - 2012b],
so that the small cells can be considered as part of mobile operator network and offload
the macro cell traffic.

13.3.8 Techno-economic Module

The feasibility of any network deployment projects can only be decided using a
techno-economic analysis that also includes cash flow (CF) and NPV considerations.
By referring only to OPEX and CAPEX it is possible to understand the costs aspects
of the project, but it is not possible to assess the feasibility of the project from the cost
perspective. Cash flow refers to the amount of money gained (i.e. revenue) and spent
during the lifetime of the network. After calculating the cash flow, it is possible to
estimate the NPV. Equation 13.3 indicates the extra gain that a project will bring with
respect to the invested money [see Nikolikj and Janevski - 2015]. If the NPV is negative,
the project will be typically rejected as it does not have any financial benefit.
∑
Ln
CFj
NPV = (13.3)
j=1
(1 + r)j
where Ln , CFj , and r, denote the network lifetime, cash flow in a year j, and the discount
rate, respectively. The “discount rate is a factor for estimating the present value of the
future cash flows by considering the time value of the money and the risk or uncertain-
ties of future incomes.
284 13 Techno-economic and Business Feasibility Analysis of 5G Transport Networks

13.4 Case Study

This section presents a case study where the proposed business feasibility framework is
applied. In the case study, we calculate the overall cost to deploy and operate a mobile
network including both the backhaul and the RAN segment considering a network life-
time of 10 years.

13.4.1 Application of Methodology/Scenarios

Topology model A 5 × 5 km dense urban area representing an average European city
with a population density of 3000 users per kilometer [see Auer et al. - 2012] is con-
sidered. The area consists of 100 multistory buildings per square kilometer, with five
floors per building and two apartments per floor. The buildings are placed according
to the Manhattan model [see Marsan et al. - 1991], which is a geometric model widely
considered in dense urban areas.

Architecture model Two options are assumed for the wireless deployment: homogeneous
(i.e. using macro BSs only), and heterogeneous (i.e. macro BSs serve outdoor users while
small cells are deployed indoors to provide coverage inside for indoor users). The guar-
anteed bandwidths of 300 Mbps per building and 600 Mbps per macro cell are assumed.
Two backhaul technologies are considered in the case study: microwave and fiber. For
the microwave only case, P2P microwave links are used to backhaul the data traffic from
both the macro BS and the small cells. At each building, a switch gathers the data traffic
from all its indoor cells and send them via a rooftop microwave antenna to the closest
hub in the area. The hubs then transfer the data traffic to the metro/aggregation network
via other P2P microwave links. Traffic from the macro cells is also sent to the backbone
via one or more hops of P2P microwave links depending on the distances from the metro
node (Figure 13.6).
In the scenario with fiber backhauling, the data traffic from the indoor users is col-
lected using a fiber to the building architecture, i.e. an aggregation switch inside each

Macro BS

MW antenna

MW hub

Fiber link

Mw link

Metro Network

Figure 13.6 Considered microwave based backhaul architecture.

 13.4 Case Study 285

Macro BS

OLT

Splitter

AWG

Fiber link

ONU
Metro Network

Figure 13.7 Considered fiber based backhaul architecture.

building is co-located with an ONU in the building basement, connected to an OLT

using a PON architecture (Figure 13.7). The macro cells are also backhauled following
the same way (i.e. one ONU per macro site connects to the OLT in the central office
located in the corner of the considered area via an optical distribution network). Since
the infrastructure deployment cost is known as the most expensive process of the TCO
in case of fiber networks, two scenarios are assessed for the fiber-based backhaul. If
no fiber infrastructure is available in the area, the mobile operator needs to deploy (i.e.
trench) its infrastructure. However, if there is already an existing fiber infrastructure
available, the operator could lease fiber instead of trenching. It is assumed that the fiber
backhaul is based on the hybrid time and wavelength division multiplexing PON con-
cept promoted by the full service access network forum as the candidate technology for
the future high capacity optical access networks [see FSAN Group - 2018].
In this technology, a four wavelength channel at 10 Gbps is sent from the transceiver
array in the OLT to a first remote node where an array wavelength grating routes
each wavelength to a separate splitter located in a second remote node [see Mahloo
et al. - 2014a]. Each wavelength coming to a splitter is shared among all the connected
ONUs using time division multiplexing technology (Figure 13.7). Based on the assump-
tion mentioned above, the following six scenarios are considered for the case study:
• Scenario 1: homogeneous RAN deployment backhauled via microwave links
(Ho_MW).
• Scenario 2: microwave backhaul is used to serve a HetNet for RAN (He_MW).
• Scenario 3: operator deploys its fiber infrastructure (i.e. trenching is required) to
backhaul its homogeneous RAN network (Ho_Tr).
• Scenario 4: operator leases fiber to backhaul homogeneous RAN network (Ho_Le).
• Scenario 5: operator deploys its fiber infrastructure (i.e. trenching is required) to
backhaul its HetNet for RAN (He_Tr).
• Scenario 6: operator leases fiber to backhaul its HetNet for RAN (He_Le).

Market model The main criteria considered for the dimensioning of RAN is the required
throughput per square kilometer in each year (shown in Table 13.1) as well as the
286 13 Techno-economic and Business Feasibility Analysis of 5G Transport Networks

Table 13.1 Considered throughput per square kilometer [see Tombaz et al. - 2014].

Year 2014 2016 2018 2020 2022 2024

Throughput (Mbps) 15 30 60 119 235 470

80%
Customer penetration

60%

40%

20%

0%
2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024
Year

Figure 13.8 User penetration curve.

coverage constraint, which is fulfilled using macro and small cell deployments. We
used the model presented in [Tombaz et al. - 2014] to calculate the required number of
macro and small cells. Figure 13.8 represents the considered penetration curve related
to the mobile customers joining the network in each year.
Table 13.2 summarizes the cost parameters used to calculate the TCO. The fault man-
agement cost is calculated based on the values in [see Mahloo et al. - 2013; Telecom
India - 2014]. The results of the dimensioning work for the RAN, based on the inputs
from the market module, are shown in Table 13.3. We assume to have one small cell per
building in the first year in order to have enough indoor coverage. Half of the users are
covered using the macro, and the other half is served by the indoor cells.

Business model In order to assess the total required investment cost of an MNO
in the worst case scenario, we selected the business model number 3 presented in
section 13.3.7. For small cells, we consider the open subscriber group model where
indoor cells are managed and owned by the MNO. We assume that the operator has
30% of the market share in the region, i.e. only 30% of the total users are considered for
the revenue calculations.

13.4.2 Techno-economic Evaluation Results

The results of the techno-economic evaluation of the scenarios as mentioned above are
presented in this section. Figure 13.9 shows the total cost of ownership for the mobile
network considering the TCO values for the backhaul and the RAN during ten years of
network lifetime. It is evident that HetNet deployments decrease the cost of the RAN
dramatically compared to the homogeneous case. However, according to the results, the
 13.4 Case Study 287

Table 13.2 Input values used for cost calculation gathered from
[Frias and Pérez - 2012a; Ahmed et al. - 2013; Mahloo et al. - 2014a,
2013; Paolini - 2011; Oughton and Frias - 2017].

Component/parameter Value

Number of team (𝛽) 1

Cost change factor (salary) (𝛼) 7%
Cost change factor (hardware) (𝛼) −3%
Discount rate (r) 10%
Subscription fee (𝜆) €30
Number of tech./team (Techte ) 2
Technician salary/hour (Techsj ) €52
Energy cost/kWh €0.1
Indoor yearly rental fee/m2 €220
Outdoor yearly rental fee/m2 €180
Small/large microwave antenna €500∕2000
G-ethernet switch €1800
Microwave hub + installation €20000
Ethernet switch €150
Yearly spectrum leasing/MHz €5
OLT (4x10G array transceiver) €7000
ONU €150
Power splitter (1:16/1:32) €170∕340
Fiber/km €80
Trenching/km €45000
Leasing upfront fee/km €800
Yearly fiber leasing fee/km €200
Macro base station and cell site €48000
Small indoor base station €250

Table 13.3 Number of installed cells.

Heterogeneous Homogeneous

Year Macro cell Small cell Macro cell

2014 4 2500 8
2016 6 4000 15
2018 9 5750 30
2020 18 7500 60
2022 36 10000 119
2024 72 12500 237
288 13 Techno-economic and Business Feasibility Analysis of 5G Transport Networks

14
BH_CAPEX BH_OPEX wireless
12

10
TCO (million €)

0
Ho_MW He_MW Ho_Tr Ho_Tr Ho_Le He_Le

Figure 13.9 Total cost of ownership of a mobile network including both RAN and backhaul.

backhaul for HetNet, in all scenarios, is more than twice as expensive (in terms of TCO)
than the corresponding homogeneous case. The results show that HetNet deployment
with microwave backhauling is the most expensive scenario. This is due to the com-
ponent cost and the power consumed by the microwave links, which increase almost
linearly with the number of small cells. Moreover, the bigger the area, the higher the
possibility to share some infrastructure using a fiber based backhaul compared to the
microwave backhaul case. Therefore, fiber based backhauling is more cost efficient, in
areas with a high density of small cells, even if an operator needs to deploy its fiber
infrastructure. This confirms the claim made earlier in the chapter that it is important
to carefully choose a proper backhaul technology in order to minimize the impact on
the TCO of a HetNet deployment.
Figure 13.10 presents the cost breakdown to assess the impact of each cost element
of the TCO for the backhaul segment. The considered cost elements are: fault man-
agement (FM), floor space (FS), spectrum and fiber leasing (Sp&Le), maintenance (M),
energy (En.), infrastructure (Infra.), and component cost (Equip.). From Figure 13.10 it
becomes evident that each cost item has a different impact on the TCO depending on the

FS FM M Sp&Le En. Infra. Equip.

100%
90%
Backhaul TCO (million €)

80%
70%
60%
50%
40%
30%
20%
10%
0%
Ho_MW He_MW Ho_Tr He_Tr Ho_Le He_Le

Figure 13.10 Cost breakdown of the backhaul TCO elements.

 13.4 Case Study 289

deployed strategy. For example, energy and equipment cost are the dominant part of a
microwave based backhaul while these two elements always amount to less than 30% for
a fiber based backhaul. The figure also shows that the impact of FM cost increases in the
case of HetNet deployments due to the higher amount of equipment and infrastructure
needed in the network.
Figure 13.11 shows the TCO evolution for all the scenarios presenting a TCO break-
down year by year. As can be seen from the figure both the backhaul technology and
the RAN deployment strategy influence the distribution of expenses in the time domain
during the considered ten years. In the case of HetNet deployments, huge upfront invest-
ment is required in the first year to have good indoor coverage backhauling all the small
cells (one small cell per building in the first year is considered). For the scenarios with
fiber trenching, also a significant amount of money needs to be invested in the first year
as most of the trenching needs to be done in the beginning.
Another interesting aspect presented in Figure 13.11 is the proportion of CAPEX
and OPEX for the backhaul technologies. In the scenario with microwave backhaul for
homogeneous wireless deployment, the OPEX is huge, and it increases considerably
with capacity growth. However, the OPEX is a tiny portion of the yearly expenses in
case of fiber based backhaul with trenching, where the infrastructure cost is dominant.
Considering an average monthly subscription fee of €30 per user (for voice and data)
and a discount rate of 10%, the NPV has been calculated for all the scenarios, and it is
presented in Figure 13.12. Except for the case of HetNet deployments with microwave
backhaul (i.e. where the backhaul TCO is extremely expensive), all the scenarios have
a positive NPV and can be considered economically viable. Another interesting aspect
to notice is the following. The He-Tr deployment has the lowest TCO value compared
to all three homogeneous scenarios. On the other hand, its NPV (i.e. the total amount
of profit at the end of the 10 year network lifetime) is the lowest. This is because most
of the investment for both backhaul and RAN needs to be done in the first years of
the project. Typically the same amount of money is worth more at present than in the
future due to the potential of producing income if invested earlier. This example shows
the importance of business viability analyses, i.e. the technology with the lowest TCO
value might not be the one economically preferable for a long-term investment project.

13.4.3 Sensitivity Analysis

There are lots of uncertainties in the input parameters considered in the cost study that
might change the results dramatically. A thorough sensitivity analysis helps to under-
stand the impact of variations in the input values on the cost results and identify the key
cost factors, which give a better view of the possible risks of the project and the reliability
of the results.
This section analyses how the variation of some key input parameters of the techno-
economic tool might impact the conclusions that were just discussed. Based on the back-
haul cost breakdown presented in Figure 13.10, we first identify the most costly elements
of each deployment option and then we calculate how the TCO of the backhaul varies
when the value of these key elements is changed. In the case of a microwave based back-
haul, power consumption and equipment costs are the most expensive elements, while
for the fiber scenarios, the infrastructure cost related to the trenching or leasing costs
has the highest share in the TCO of the backhaul. To illustrate the impact of each one
 Cost (million €)

(d)
(a)
Cost (million €)

0
1
2
3
4
5
6
0
0.4
0.8
1.2
1.6
2
2.4
2.8
3.2
20 20
14 14
20 20
15 15
20 20
16 16
20 20
17 17
20 20
18 18
20 20

OPEX
19 19

OPEX
20 20
20 20
20 20
21 21

CAPEX
20 20

CAPEX
2 2
20 2 20 2
23 23
20 20
24 24

Cost (million €) Cost (million €)

(e)
(b)

0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0
0.2
0.4
0.6
0.8
1
1.2
1.4
20 20
14 14
20 20
15 15
20 20
16 16
20 20
17 17
20 20
18 18
20 20

OPEX
19 19
OPEX

20 20
20 20
20 20
21 21

CAPEX
20 20
CAPEX

2 2
20 2 20 2
23 23
20 20
24 24

Cost (million €) Cost (million €)

(f)
(c)

0
0.3
0.6
0.9
1.2
1.5
1.8
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35

20 20
14 14
20 20
15 15
20 20
16 16

Figure 13.11 Backhaul TCO evolution per year for (a) Ho-MW, (b) Ho-Tr, (c) Ho-Le, (d) He-MW, (e) He-Tr, and (f ) He-Le scenarios.
20 20
17 17
20 20
18 18
OPEX

20 20
OPEX

19 19
20 20
20 20
20 20
CAPEX

21 21
CAPEX

20 20
2 22
20 2 20
23 23
20 20
24 24
 13.4 Case Study 291

3
NPV (million €)

0
Ho_MW He_MW Ho_Tr Ho_Le He_Tr He_Le
–1

–2

–3
–4

Figure 13.12 NPV at the end of the 10 year network lifetime.

24

22

20
TCO (million €)

18

16

14

12

10

8
Ho_MWHe_MW Ho_MWHe_MW Ho_Tr He_Tr Ho_Le He_Le

(a) Energy (b) Antenna (c) Trench (d) Leas

Figure 13.13 Sensitivity analysis of the TCO varying: (a) energy cost (±50%), (b) MW antenna price
(±50%), (c) amount of re-usable trenching (from 0% to 100%), (d) price of leasing (±50%).

of these elements, Figure 13.13 shows the fluctuation of the TCO of a mobile network
deployment (including both RAN and backhaul).
The grey circles in Figure 13.13, represent the TCO values for each scenario calculated
in the previous section, and the green bars demonstrate its variation if the values of cor-
responding input parameters are changed. It can be seen that even by decreasing the
energy price by half or by using antennas costing half the price, microwave based solu-
tions are still the most expensive options among all the scenarios. An interesting finding
from Figure 13.13 is the difference between a greenfield scenario where no infrastructure
is available (0%) and a brownfield one where a certain amount of fiber infrastructure is
available, and only partial trenching is needed (> 0%). If more than half of the required
trenching is available, then fiber based backhaul for a HetNet scenario offers the low-
est TCO for providing mobile services. The price of leasing fibers, which varies a lot
292 13 Techno-economic and Business Feasibility Analysis of 5G Transport Networks

20
Ho_MW He_MW Ho_Tr Ho_Le He_Tr He_Le
15

10
NPV (million €)

0
10 15 20 25 30 35 40 45 50

–5

–10

–15

Figure 13.14 Sensitivity analysis of NPV based on the variation of the monthly subscription fee.

depending on the country and region, has a huge influence on the investment cost. With
a 50% increase in the leasing price, fiber trenching becomes a cheaper and more attrac-
tive option than fiber leasing from a TCO point of view.
Another important parameter in any business viability analyses is the revenue that is
related to the project profitability. In order to investigate the impact of revenue per user
and to find the lowest monthly subscription fee resulting in a positive NPV, Figure 13.14
presents a sensitivity analysis based on the variation of the subscription fee per user. It
is evident that if the average monthly fee per user is more than €40 all six scenarios,
even He-Mw, can reach a positive NPV, and all the cases have a negative NPV when the
charged monthly fee is €15 or less.

13.5 Conclusion
In summary, this chapter presented a comprehensive techno-economic framework for
estimating the TCO of a backhaul network segment as well as for analyzing the business
viability of a given wireless network deployment. The chapter focused on two backhaul
technologies: microwave and fiber. The proposed framework is applied to a case study to
estimate the TCO and the NPV of a heterogeneous and homogeneous wireless deploy-
ment. The results showed a considerable increase in the value of the backhaul TCO
in the case of a heterogeneous deployment compared to a conventional homogeneous
scenario. The results also showed that fiber is the most cost-efficient technology to pro-
vide a high capacity backhaul for heterogeneous wireless deployments. The cheapest
alternative is to lease fiber connectivity when possible. Our results also highlight the
importance of selecting the right backhaul technology in order to keep the economic
benefits brought by heterogeneous wireless deployments. The proposed business via-
bility analysis also indicated the importance of having a complete techno-economic
assessment instead of having a pure cost calculation. Our NPV results showed that a
lower TCO does not always lead to a higher profit because in the long-term project
investments done in different time periods affect the total profit of the project.
 Bibliography 293

Bibliography

A. A. W. Ahmed, J. Markendahl, C. Cavdar, and A. Ghanbari. Study on the effects of

backhaul solutions on indoor mobile deployment macrocell vs. femtocell. Proc. of IEEE
International conference on Personal, Indoor and Mobile Radio Communications
(PIMRC), 2013.
M. Aleksic, S. and Deruyck, W. Vereecken, W. Joseph, M. Pickavet, and L. Martens. Energy
efficiency of femtocell deployment in combined wireless/optical access networks.
Elsevier Computer Networks, 57:1217–1233, 2013.
Gunther Auer, Oliver Blume, Vito Giannini, Istvan Godor, Muhammad Ali Imran, Ylva
Jading, Efstathios Katranaras, Magnus Olsson (EAB), Dario Sabella, Per Skillermark, and
Wieslawa Wajda. Energy efficiency analysis of the reference systems, areas of
improvements and target breakdown. EARTH Deliverable D2.3, 2012.
I. Chih-Lin, C. Rowell, S. Han, Z. Xu, G. Li, and Z. Pan. Toward green and soft: A 5G
perspective. IEEE Communications Magazine, 52:66–73, 2014.
C. Cid, M. Ruiz, L. Velasco, and G. Junyent. Costs and revenues models for optical
networks architectures comparison. IX Workshop in G/MPLS Networks, 2010.
Cisco. Cisco visual networking index: Global mobile data traffic forecast update 2013:2018.
Cisco Inc., USA, C11-738429-00, 2015.
H. Claussen, L.T.W. Ho, and F. Pivit. Effects of joint macrocell and residential picocell
deployment on the network energy efficiency. Proc. of IEEE International Symposium on
Personal, Indoor and Mobile Radio Communications (PIMRC), 2008.
J. Coldrey, M. Berg, L. Manholm, C. Larsson, and J. Hansryd. Non-line-of-sight small cell
backhauling using microwave technology. IEEE Communications Magazine, 51:78–84,
2013.
CPRI. Common public radio interface (CPRI), v6. 0. CPRI Specification, 6, 2013.
Ercisson AB. Ericsson microwave towards 2020 report. Ercisson AB, 2014.
Ercisson AB. Ericsson microwave outlook. Ercisson AB, 13:97–113, 2017.
Ericsson AB. Ericsson radio dot system. Ericsson AB, white paper, 2013.
F. S Farias, P. Monti, A. Vastberg, M. Nilson, J. C. W. A. Costa, and L. Wosinska. Green
backhauling for heterogeneous mobile access networks: What are the challenges? Proc.
of International Conference on Information, Communications and Signal Processing
(ICICS), 2013a.
F. S. Farias, M. Fiorani, S. Tombaz, M. Mahloo, L. Wosinska, J. C. W. A. Costa, and P. Monti.
Cost-and energy-efficient backhaul options for heterogeneous mobile network
deployments. Photonic Network Communications, 32(3): 422–437, 2016.
F.S. Farias, Borges G.S., Rodrigues R. M, A. L. Santana, and J.C.W.A. Costa. Real-time noise
identification in dsl systems using computational intelligence algorithms. Proc. of
International Conference on Advanced Technologies for Communications (ATC), pages
252–255, 2013b.
Femto forum. Femtocells-natural solution for offload. Femto forum white paper, 2010.
D. Feng, W. Sun, and W. Hu. Hybrid radio frequency and free space optical communication
for 5G backhaul. 2014.
Z. Frias and J. Pérez. Techno-economic analysis of femtocell deployment in long-term
evolution networks. EURASIP Journal on Wireless Communications and Networking,
2012:1–12, 2012a.
294 13 Techno-economic and Business Feasibility Analysis of 5G Transport Networks

Z. Frias and J. Pérez. Techno-economic analysis of femtocell deployment in long-term

evolution networks. EURASIP Journal on Wireless Communications and Networking,
2012:1–12, 2012b.
FSAN Group. Full service access network. http://www.fsan.org, 2018.
T. Geitner. Vodafone group technology update. Vodafane, 2005.
T. Giles, J. Markendahl, J. Zander, P. Zetterberg, P. Karlsson, G. Malmgren, and J. Nilsson.
Cost drivers and deployment scenarios for future broadband wireless networks - key
research problems and directions for research. Proc. of IEEE Vehicle Technology
Conference (VTC), 2004.
Juniper Networks. Mobile backhaul reference architecture. Juniper Networks white paper,
2011.
J.S. Khirallah, C. and Thompson and H. Rashvand. Energy and cost impacts of relay and
femtocell deployments in long-term-evolution advanced. IET Communications,
5:2617–2628, 2011.
F-C. Kuo, F.A. Zdarsky, J. Lessmann, and S. Schmid. Cost efficient wireless mobile backhaul
topologies: an analytical study,. Proc of IEEE Global Telecommunications Conference
(GLOBECOM), 2010.
S. Lins, P. Figueiredo, and A. Klautau. Requirements and evaluation of copper-based mobile
backhaul for small cells LTE networks. Proc. of IEEE International Microwave and
Optoelectronics Conference (IMOC), 2013.
M. Mahloo, C. M. Machuca, J. Chen, and L. Wosinska. Protection cost evaluation of
wdm-based next generation optical access networks,. Journal of Optical Switching and
Networking (OSN), 10:89–99, 2013.
M. Mahloo, J. Chen, L. Wosinska, A. Dixit, B. Lannoo, D. Colle, and C. M. Machuca.
Toward reliable hybrid WDM/TDM passive optical networks. IEEE Communication
Magazine, 52:14–23, 2014a.
M. Mahloo, P. Monti, J. Chen, and L. Wosinska. Cost modeling of backhaul for mobile
networks. Proc. of IEEE International Conference on Communications (ICC), 2014b.
J. Markendahl and Ö. Mäkitalo. A comparative study of deployment options, capacity and
cost structure for macrocellular and femtocell networks. Proc. of IEEE International
Symposium on Personal Indoor and Mobile Radio Communications (PIMRC), 2010.
J. Markendahl, Ö. Mäkitalo, and J. Werding. Analysis of cost structure and business model
options for wireless access provisioning using femtocell solutions. Proc. of European
International Telecommunications Society (ITS) Conference, 2008.
M. A. Marsan, G. Albertengo, A. Francese, and F. Neri. Manhattan topologies for passive
all-optical networks. Proc. of Annual European Fiber Optic Communications and Local
Area Network Exposition, 1991.
P. Monti, S. Tombaz, L. Wosinska, and J. Zander. Mobile backhaul in heterogeneous
network deployments: technology options and power consumption. Proc. of
International Conference on Transparent Optical Networks (ICTON), 2012.
S. Nie, G. R. MacCartney, S. Sun, and T. S Rappaport. 72 GHz millimeter wave indoor
measurements for wireless and backhaul communications,. Proc. of International
Symposium on Personal Indoor and Mobile Radio Communications (PIMRC), 2013.
V. Nikolikj and T. Janevski. State-of-the-art business performance evaluation of the
advanced wireless heterogeneous networks to be deployed for the “tera age”. Wireless
Personal Communications, 84(3):2241–2270, 2015.
Nokia. G.fast. https://networks.nokia.com/solutions/g.fast, 2018.
 Bibliography 295

T. Norman. Wireless network traffic 2010-2015: forecasts and analysis. Analysys Mason,
2010.
A. Osseiran et. al. The foundation of the mobile and wireless communications system for
2020 and beyond: Challenges, enablers and technology solutions. Proc. Vehicular
Technology Conference (VTC Spring), pages 1–5, 2013.
E. J. Oughton and Z. Frias. The cost, coverage and rollout implications of 5g infrastructure
in britain. Telecommunications Policy, 2017.
M. Paolini. An analysis of the total cost of ownership of point-to-point, point-to-
multipoint, and fibre options. White paper on crucial economics for mobile data
backhaul, 2011.
Senza fili. Crucial economics for mobile data backhaul. Senza fili consulting white paper,
2011.
Skyfiber. How to meet your backhaul capacity needs while maximizing revenue. Skyfiber
white paper, 2013.
W-S. Soh, Z. Antoniou, and Hyong S. Kim. Improving restorability in radio access
networks. Proc of IEEE Global Telecommunications Conference (GLOBECOM), 2003.
Telecom India. Consultation paper on allocation and pricing of microwave access (mwa)
and microwave backbone (mwb) rf carriers. Telecom regulatory authority of India, 2014.
O. Tipmongkolsilp, S. Zaghloul, and A. Jukan. The evolution of cellular backhaul
technologies: Current issues and future trends. IEEE Communications Surveys &
Tutorials, 13:97–113, 2011.
S. Tombaz, P. Monti, K. Wang, A. Vastberg, M. Forzati, and J. Zander. Impact of
backhauling power consumption on the deployment of heterogeneous mobile networks.
Proc. of IEEE Global Telecommunications Conference (GLOBECOM), 2011.
S. Tombaz, P. Monti, F. Farias, M. Fiorani, L. Wosinska, and J. Zander. Is backhaul
becoming a bottleneck for green wireless access networks? Proc. of IEEE International
Conference on Communications (ICC), 2014.
UMTS Forum. Mobile traffic forecasts 2010-2020. UMTS Forum, Nokia, 2011.
S. Verbrugge, K. Casier, J. Van Ooteghem, and B. Lannoo. white paper: Practical steps in
techno-economic evaluation of network deployment planning. 2009.
F. Yaghoubi, M. Mahloo, L. Wosinska, P. Monti, F. S. Farias, and J. C. W. A. Costa. A
techno-economic framework for 5G transport networks. IEEE Wireless
Communications, 25: 56–63, 2018.
J. Zander and P. Mähönen. Riding the data tsunami in the cloud: myths and challenges in
future wireless access. IEEE Communications Magazine, 51: 145–151, 2013.