Nº 1 2013 > Future networks
How ITU can help develop future networks
Martin Waldburger, from the University of Zurich, and Tatiana Kurakova, an ITU-T Study Group 13 Counsellor
Standardization demands resources, and in particular the precious time of experts, so it makes sense to direct standardization efforts towards future network technologies that are most likely to be successful. But how is it possible to predict the emerging winners? Study Group 13 of ITU’s Telecommunication Standardization Sector (ITU–T) has been thinking about this and proposes some approaches in this article.
How will future networks be formed? Which technological innovations will thrive? The answers are impossible to consistently and reliably predict. Yet ITU–T needs to develop standards that will lead the world to seamless global communications, ensuring interoperability of future networks.
ITU–T can take two approaches to future network standardization. First, it can gather its members and outline high-level objectives and design goals for future networks. Second, it can help technology designers and standardization bodies anticipate whether or not a future network technology is likely to be successful.
ITU–T Study Group 13, and especially its Question 21 group on “Future Networks”, is working on both these approaches. The Question 21 group has successfully addressed the first by finalizing ITU–T Recommendation Y.3001, which identifies objectives and design goals for future networks. The group is now working on the second approach, focusing on methods to anticipate whether or not a future network technology is likely to be successful.
Which network technologies will be successful?
The key factor in predicting the adoption potential of future network technology is to assess whether the design of the technology takes socio-economic impact factors into account — in other words, whether the technology is designed with an awareness of the potential for vigorous struggles between conflicting interests (tussles).
Many technologies have failed to be deployed, to flourish, or to be sustainable, because of inadequate or inappropriate decisions by the architect or because of network limitations. The designs of failed technologies did not take account of the important socio-economic impacts that the technology would have on stakeholders.
With this in mind, the Question 21 group began in 2012 to collect, investigate, and recommend, useful methods for designing technology, with socio-economic awareness, and in an incentives-compatible manner. This resulted in a draft recommendation, known provisionally as ITU–T Recommendation Y.FNsocioeconomic, which proposes tussle analysis as a method to address the socio-economic aspects of future network technology.
When a candidate technology is put forward, tussle analysis allows for:
- assessment, at the technology design phase, of the socio-economic impact of the technology, taking into account the relevant set of stakeholders, tussles among them, and the range of available choices;
- determining whether a stable and incentives-compatible outcome, or an unstable outcome, will result from deploying the technology;
- identification of any potentially unwanted spillover effects from the technology’s primary functionality to another functionality.
Identifying stakeholders and their tussles
Tussle analysis consists of three main steps: step 1, identifying stakeholders; step 2, identifying tussles among these stakeholders; and step 3, assessing the impact and evolution of tussles. The draft recommendation documents these three steps and describes suitable methods to implement each one of them.
Step 1 involves identifying all primary stakeholders and their characteristics for the functionality under investigation. It is important to identify as complete a set of stakeholders as possible, since all further steps in tussle analysis depend on the range of identified stakeholders. If a stakeholder is missing from the list, potential tussles involving that stakeholder cannot be captured in step 2, rendering the analysis incomplete, or leading to incorrect results in step 3.
Personal observation, interviews, and to a certain extent, role-playing simulations, are recommended methods for identifying a complete and relevant set of stakeholders.
Any person knowledgeable enough about the characteristics of future network technology is assumed to be able — from personal observation — to determine an a priori complete and relevant set of candidate stakeholder roles. Personal observation in this context is understood to cover not only personal experience and knowledge, but also study of the relevant literature.
Personal observation and interview methods are especially effective when combined. While personal observation is recommended to determine an a priori candidate stakeholder role list, interviews can be used to validate a stakeholder’s involvement (relevance) and to identify other relevant stakeholders that were not incuded in the a priori stakeholder role list (completeness). Role-playing simulation methods (for example, the Delphi method and focus groups) are a recommended alternative provided that participants are selected to be representative.
Step 2 in the tussle analysis consists of identifying tussles among identified stakeholders. Step 2 depends on a successful step 1 (stakeholder identification) and it provides input to step 3 (tussle impact and evolution). Any method chosen to implement step 2 has to achieve a high level of completeness and relevance with respect to stakeholder incentives, conflicts and available choices. This is not always easy to determine because information may be considered confidential (from a business perspective), or there may be a hidden agenda.
In the context of step 2, relevance is even more important than completeness. Tussles tend to vary in terms of complexity (for example, involvement of more than two stakeholders, with incentives that are difficult to assess). It is therefore important to identify as many tussles as possible, but it is even more important to rank the tussles according to their relevance. Assessment of the impact on stakeholders (step 3) is applied only to relevant tussles.
A variety of methods are recommended for identifying a complete and relevant set of tussles. These are role-playing simulation, personal observation, risk management, the MACTOR (matrix of alliances and conflicts: tactics, objectives and recommendations) method and, to a certain extent, SWOT (strengths, weaknesses, opportunities and threats) analysis and interviews.
Assuming that participants are selected carefully, the most recommended method is role-playing simulation (such as the Delphi method or focus groups) because it is a group-based rather than an individual approach. Bringing several motivated and knowledgeable experts into a group (a focus group, for example) greatly increases the likelihood of relevant tussles being identified through confrontational, debate-oriented, and direct (but moderated) interaction among participants.
Individual approaches, such as personal observation, risk management and the MACTOR method are also recommended, especially when personal observation is combined with risk management or with the MACTOR method. The focus of risk management on identifying factors that may have a negative effect on a system is very helpful in identifying tussles, whether the effect is evaluated quantitatively or qualitatively, and even if the evaluation is based on the opinions of individuals rather than of a group of experts. Equally, the MACTOR method can be used to identify tussles thanks to its capacity for giving an overview of possible alliances and conflicts in a business ecosystem.
Assessing the impact of technologies on stakeholders
Tussle analysis is about anticipating the impact of a candidate technology (and its anticipated functionality) by considering the potential tussles that may emerge among stakeholders.
Step 3 in tussle analysis consists of assessing, for each identified tussle, the impact for each stakeholder and any potential spillovers. Anticipation of tussles needs to be based on reasonable and traceable grounds. In contrast to steps 1 and 2, where the goals are completeness and relevance, in step 3 the focus is on ensuring that each and every step in the predicted evolution of a tussle is plausible. The best methods are formal (producing reproducible results), risk-oriented (dealing with uncertainty and probability), and dynamic (comprising complex system modelling with feedback cycles). The recommended methods for assessing tussle impact and evolution are game theory, risk management, system dynamics, role-playing simulations and, to a certain extent, interviews, the MACTOR method and SWOT analysis. Appendix I of ITU–T Recommendation Y.FNsocioeconomic provides an overview of these methods.
Game theory is the recommended method when a single tussle with well-known incentive structures can be modelled and played out in a formal and quantified manner. Where strategy candidates are available, game theory modelling is best adopted for tussles that derive from multi-party interactions. Risk management allows for a qualitative and quantitative evaluation of previously identified factors. Game theory and risk management are, however, less suitable for assessing evolution over time.
System dynamics is the best method to cope with simulations of various outcomes when multiple stakeholders interact over a longer modelling horizon. By focusing on actions and counter-actions in an interactive, complex system, system dynamics is the primary recommended method for step 3 in tussle analysis — with the trade-off that it needs a major modelling effort.
Where a non-quantified method is acceptable, role-playing simulations, in particular focus groups, are recommended. By providing an opportunity to observe stakeholder behaviour, focus groups offer a unique understanding of the dynamics of major tussles.
Interviews, the MACTOR method, and SWOT analysis may also be valid complementary instruments for validating results originating from the use of the other recommended methods.
Looking ahead to a standard on future network technology
When ITU–T Recommendation Y.FNsocioeconomic is finalized — most likely during 2013 — its recommendations on the socio-economically aware design of future network technology will become an official ITU–T standard. What will be the consequences of these recommendations, and how will they be implemented?
All the stakeholders in ITU–T, which brings together administrations, industry players (such as equipment manufacturers or network operators) and academia, or in ITU’s Telecommunication Standardization Bureau (TSB), which provides ITU–T with logistical and operational support, might be affected in different ways by the recommendations. For instance, because any standardization organization has limited resources, it makes sense to prioritize stardardization activities. In the case of ITU–T, the various hierarchical levels — question groups, working parties, and study groups — might apply recommended socio-economic measures as a strategic instrument to decide which technologies to standardize.
If the new recommendations on future network technology incorporate a section on socio-economic assessment of the standardized technology, TSB could prepare the necessary template documents, and support rapporteurs and editors in organizing meetings to produce such assessments. For example, TSB might help by setting up a focus group, or by providing training.
In current and future work related to the socio-economically aware design of future network technology, the Question 21 group (and its new Question 16/13 on “Environmental and socio-economic sustainability in future networks and early realization of FN” in the study period 2013–2016) will have to identify the type of cases for which the new ITU–T Recommendation Y.FNsocioeconomic would be useful, whether within the question group, working party, study group, or even ITU–T as a whole. There are many risks and uncertainties associated with new technologies, but one certainty is that there is a need for ITU to help technology designers and standardization bodies anticipate whether or not a future network technology is likely to be successful.