You are here: Project > Technical Highlights > Scenarios and Business Models > 
27.3.2017 : 12:31

Highlights: Scenarios and Business Models

The QoSMOS projct has defined scenarios and system requirements, an ecosystem and metrics for spectrum micro-trading and performed a business case analysis for opportunistic access based on cognitive radio in the TV whitespaces.

QoSMOS Scenarios

The scenarios describe the context in which the QoSMOS system is believed to form a feasible solution. The project has defined six distinct scenarios for opportunistic access enabled by cognitive radio used as the basis for the technical work.  The scenarios provide a good mix between long and short range as well as cellular and non-cellular architecture.

Dynamic backhaul

  • Wireless backhaul connections between access networks and remote terminals and a core network where the nodes could be several point-to-point links (relays).

Cellular extension in whitespace

  • Mobile networks (e.g. LTE) can utilise whitespace (WS) spectrum in addition to their own licensed spectrum. This additional spectrum allows for mobile operators to gain additional bandwidths to benefit the user, or additional coverage using low frequency WS.

Rural broadband

  • Wireless Internet connectivity to homes in rural locations can utilize WS on low frequencies. The homes may be 1-10km from the base station.

Cognitive ad hoc network

  • Ad hoc networks may include one or more nodes with access to the Internet via other networks, but it may possibly be completely stand alone. Ad hoc networks could be used, for example, for creating communications networks for first responders to emergency situations.

Direct terminal-to-terminal in cellular

  • In an infrastructure-based network, the mobile terminals can communicate directly between each other, with no data traffic going through the base station. The system management is still within the cellular network rather than in the terminals themselves. This can save the resource of the base station and can reduce power consumption in mobile terminals.

Cognitive femtocell

  • A user situation with low mobility, but high demands on throughput and QoS. Femtocells are always connected to an infrastructure. Femtocell are typically used as domestic wireless broadband solutions and public hot spots in e.g. commuter areas, cafés and similar. Both indoor and outdoor deployment is possible.

The scenarios and rationalization behind them are all comprehensively described in the QoSMOS deliverable D1.2 from December 2010. A subset of these scenarios are further used in studies on business models, technical analysis and proof-of-concepts.


The QoSMOS scenarios

QoSMOS System Requirements

The requirements for the QoSMOS system have been defined according to four major goals:

  • Competitiveness
  • Regulatory compliance
  • Technical performance
  • Flexibility and scalability
QoSMOS frequency flexibility

Frequency flexibility

The QoSMOS approach to requirements also contains the concept of frequency flexibility. Different parts of the QoSMOS system can be placed on a scale from no frequency dependence to high frequency dependence. 

In order to meet the overall goals, four top-level requirements have been defined:

Requirements for business, user and service

  • The QoSMOS system should be competitive to other technologies and show a proven benefit in relevant markets and scenarios. (Competitiveness of the QoSMOS system)

Requirements for system operation

  • The QoSMOS system shall be flexible and adaptable to differences in regulations given for the regions and markets in which it is intended to be deployed. (Regulatory compliance)

Requirements for performance

  • The QoSMOS system’s technical performance should be good enough to meet users’ expectations of the service delivered. (Technical performance)

Requirements for architecture and complexity

  • The QoSMOS architecture shall ensure complying with other external systems and ensure flexibility and scalability. (Architecture and complexity)

Each of these four requirements is further broken down into more specific requirements which are detailed in the QoSMOS Deliverable D1.4 from March 2011.


Roles in the spectrum trading ecosystem

Spectrum Micro-Trading

Spectrum micro-trading can be defined as the possibility to trade spectrum resources on the micro-scale in one or more of the spatial, temporal and frequency dimensions. This would enable wireless services to acquire spectrum for small or wide geographical areas, for short or long time periods and for narrow or wide bandwidths. Hence, spectrum utilization and the opportunity to acquire spectrum resources might increase when optimizing metrics and specifying policies properly. Regulatory rules for spectrum trading have been implemented in some countries for some spectrum bands, for instance in the UK and US. However, the current spectrum trading regimes usually require long times to execute a trade, hence limiting the flexibility at short time scales.

Ecosystem for Spectrum Trading

The QoSMOS project has identified the ecosystem and roles to be involved in a spectrum micro-trading market. The main parameter identified enabling trading between different actors is the information about available spectrum resources, which can be defined by its spatial, temporal and frequency dimensions. Other parameters such as maximum transmit power and regulatory constraints on spectrum resource usage are also identified.


The micro-trading pixelation model

The Spectrum Micro-Trading Pixelation Model

A model referred to as the “spectrum micro-trading pixelation” model is proposed. Using this model, spectrum micro-trading can be implemented in all dimensions, the micro-spatial, micro-temporal and micro-frequency scale. Each dimension is defined as pixels whose micro-granularity can be specified to fixed parameters for optimized performance by using the metrics defined.

Metrics for Spectrum Trading

A focused study of metrics for the evaluation of spectrum-micro trading systems and markets results in the definition of 10 key metrics. These metrics might be specific to one or more of the roles and actors defined in the ecosystem; for example, social welfare could be specific to the regulator.

Metrics for spectrum micro-trading
Liquidity An asset’s ability to be sold without causing a significant movement in the price and with minimum loss of value
Trading volume The number of completed trades
Spectrum price Low values of spectrum price would indicate that there is an excess in supply of spectrum, whereas high values would indicate that there is low supply or high demand
Profitability In telecommunication it is common to measure the NPV (Net Present Value) of a project when evaluating the profitability
Blocking ratio The number of spectrum buyers or leasers that fail to acquire spectrum through the spectrum market
Spectrum exploitation efficiency A measure of how well the allocation algorithm exploits available spectral resources, regardless of the time taken to perform the allocation
Spectrum allocation delay The time required to allocate spectrum. This will be a measure of market overhead
Interference temperature Interference is the level of electromagnetic disturbance from other RF sources and noise
User experience A subjective measure of the quality of using the spectrum micro-trading market by the user
Social welfare An important metric for the regulator when introducing spectrum micro-trading. Social welfare is a broad term meaning the well-being of the entire society
Spectrum exploitation efficiency. Each SP has 8MHz of permanent spectrum. B=8MHz (red), 4MHz (green), 2MHz (blue) and 1MHz (magenta)

Micro-Trading Market Simulations

A simulation model using agent based multi-reinforcement learning is implemented to analyze a spectrum micro-trading market and to evaluate the metrics defined. A simulation scenario is studied based on the “Cellular extension in white spaces” scenario defined in the QoSMOS project in undefinedD1.2. It is based on the Q-learning heuristics introduced by undefinedC. Watkins in 1989. It is a simple form of machine learning intended to make an agent perform state transitions which improve its long-term reward.

As an example, the spectrum exploitation efficiency metric is evaluated for the case of six service providers (SP), each having 8 MHz of permanent spectrum and bidding for extra spectrum in blocks of 1, 2, 4 or 8 MHz. The following assumptions apply:

  • User demand fluctuates on a time-scale much longer than the auction period; we studied the case in which the demand has a strong 24-hour periodicity, as typically observed in retail networks.
  • During the period the demand is varying cyclically with two high peaks illustrating the typical busy hour behaviour in the morning and evening. The auction is run periodically and synchronously.
  • The auction sells multiple units of spectrum blocks and allocation of spectrum by the broker is proportional to the bid price.

This initial simulation shows that smaller spectrum block sizes improve the spectrum allocation efficiency. Further simulations are planned for future publications.

The ecosystem, metrics and proposed trading model is further described in the undefinedQoSMOS deliverable on Spectrum Micro Trading Analysis, D1.5 from July 2012.

Business and Deployment Models Analysis and Evaluation

The targeted business case scenarios represents a set of very different types of applications where the technology would be used by different types of actors, which reflects the general applicability of the QoSMOS system.

Business Opportunities for Cognitive Radio Systems

The most interesting business opportunities for cognitive radio are

  • Easier access to new markets. By using cognitive radio and use spectrum opportunistically, the start-up time and initial costs will be significantly reduced.
  • Opportunities for new entrants. Many companies offer services that can potentially be enhanced or extended by combining them with a wireless service, but has been prevented from doing so due to the high investments required to get access to spectrum.
  • Capacity enhancements. An existing wireless operator having an infrastructure in an area can use opportunistic spectrum to increase the capacity of its network.
  • Spectrum sharing. Cognitive radio can be used as an enabling technology for spectrum sharing between operators. Cognitive radios have the ability to get information about how the spectrum is used in their environments and adapt their transmissions to minimize the disturbance to other radios.
  • Spectrum trading. Spectrum trading is a business opportunity for wireless operators to get revenues from their spectrum at times when they do not need all of it themselves. Spectrum trading is also an opportunity for new actors to enter the wireless communication ecosystem, such as spectrum brokers and pure spectrum owners.
  • Cost and performance gains in own network. Wireless operators can improve their own spectrum efficiency by using cognitive radio technology. Cognitive radio systems are able to adapt their operation according to given criteria, and can for example be used for optimizing transmissions, automating operational tasks and optimizing network tuning.
  • New services. Cognitive radios’ ability to know their environment and know their users’ needs can be used to offer environment- and context-aware type of services. Such services can be seen as an extension of location-based services, which are already offered as a service by some operators.

Cognitive radio is expected to offer business opportunities for different players, including wireless operators, fixed operators and new entrants.

Criteria for Selecting Business Cases

The business case study of QoSMOS has used the following criteria for targeting the most interesting and promising scenarios for business case studies:

  • Market Potential
  • Best Technological Solution
  • Technical Feasibility
  • Economic Feasibility
  • Regulatory Feasibility
  • Ecosystem Feasibility
  • Benefits for the society

Business Case Definitions

QoSMOS has defined and evaluated 4 different business cases:

  • Offloading of LTE networks - a cellular operator with an existing infrastructure will use TVWS spectrum in a cognitive way to enhance its mobile broadband offer. For the cellular operator, this can be seen as an alternative to waiting for some of these frequencies to be freed from TV use to mobile broadband (“Digital Dividend 2” – DD2) and acquiring spectrum there via spectrum auctions.
  • Rural broadband - a fixed-line operator provides a broadband service to rural not-spots using TVWS. The homes will have roof-top antennas pointing towards the base station.
  • Cognitive femtocell - a fixed network operator will use cognitive femtocells to extend its operation to be able to offer a mobile broadband service to its customers.
  • Machine-to-machine (M2M) in whitespace - an operator deploying the infrastructure for an M2M network in the UK. This could be an existing network operator or a new entrant.
Offloading of LTE networks
Rural broadband

Cognitive femtocell
M2M in whitespace

The business cases have been carefully evaluated from a technical point of view in carefully defined deployment environments.

The QoSMOS system can be applied to any frequency band, however in order to quantify the potential of deploying QoSMOS system solutions both technical and economic constraints must be determined. In all business cases it is assumed that the QoSMOS system operates in TV white space spectrum (470-790 MHz). The reason for this is that it is the only frequency band for which operational constraints, like the maximum allowed transmission power are given or indicated by regulators (FCC and Ofcom), and TVWS is a suitable spectrum band to use due to its favourable propagation properties as well as the fact that it is currently being opened up for opportunistic access.

Definitions of QoSMOS business cases
  Offloading of LTE networks Rural broadband Cognitive femtocell Machine-to-machine (M2M) in whitespace
Study period 2015 – 2020 2013 – 2023 2015 – 2020 2013 – 2023
Operator Cellular operator with existing LTE infrastructure Fixed-line operator Fixed line operator An operator deploying an infrastructure for an M2M network in the UK
Deployment environment A western European city with 1 million inhabitants covering an area of 200 km2 The UK with a nominal 1 million rural not-spots A western European city with 1 million inhabitants covering an area of 200 km2 The UK market
Expected range or coverage Using existing LTE grid, inter-site distance: 750 m 6 km to give > 2 Mb/s 75 – 90 m for indoor placed femtocells Ubiquitous coverage
Expected number of subscribers or terminals at end of study period 100 000 users/enabled terminals 1 000 000 users 50 000 femtocells 50 000 000 devices

Outcome of Business Case Evaluations

The business cases evaluations show a moderate to high risk. The base case NPVs vary from 3 to 115 million €, which does not reflect how relatively good the business cases are but rather the scale of the deployment considered. Since all the NPVs are positive when their base case assumptions are used, it can be concluded that all the considered business cases are potentially economic viable.

The business opportunities are discussed in more detail in the undefinedQoSMOS Whitepaper on “Business opportunities and Scenarios for Cognitive Radio”. The business case definitions and evaluations are described in detail in the QoSMOS deliverable on Economic benefits of a QoSMOS system, D1.6 from December 2012.

Business case outcome summary
  Offloading of LTE networks Rural broadband Cognitive femtocell M2M in whitespace
Study period 2015 - 2020 2013 - 2023 2015 - 2020 2013 - 2023
Base case NPV 3.0 million € better than DD2 alternative 115 million € 5.5 million € 16.4 million €
Most critical parameters Price of DD2 spectrum Subscription fee and number of subscribers Subscription fee and general OPEX costs Number of M2M devices and revenue per connected device.
Most important QoSMOS related parameters Cost of cognitive LTE terminal and cost of upgrading eNB Cost of CPE and installation Cost of cognitive femtocell, average femtocell installation costs and cognitive femtocell coverage range in suburban areas Database fee and cost of CPE
Risk Moderate High Moderate High