IEEE Talks IoT: Cicek Cavdar and Burak Kantarci

cicek cavdar burak kantarci Dr. Cicek Cavdar and Dr. Burak Kantarci are both active in the IEEE Green ICT Initiative, which is driven, in part, by the advent of the Internet of Things (IoT). Cavdar was coordinator of the project 5GrEEn: Towards Green 5G Mobile Network, now Swedish cluster coordinator of the project SooGREEN – Service Oriented Optimization of Green Mobile Networks, and a senior researcher in the Communications Systems Department at the School of ICT at KTH, the Royal Institute of Technology, Sweden. Kantarci is a senior member of IEEE, an assistant professor in the Department of Electrical and Computer Engineering at the Wallace H. Coulter School of Engineering, Clarkson University, New York, and the founding director of the Next Generation Communications and Computing Networks (NEXTCON) Lab at Clarkson. In the following Q&A they discuss the imperatives of Green ICT as IoT flourishes.

Question: How does growth of the IoT impact the urgency of the Green ICT movement?

Cavdar: Let’s begin with the drivers for Green ICT, then consider how the IoT underscores the urgency.

The carbon footprint for information and telecommunication technologies (ICT) has for some time been having global impacts. By 2007, ICT’s global CO2 footprint accounted for 2% of all emissions, roughly the same as the global aviation industry. That was a decade ago. Today, ICT’s share of global emissions is projected to double to 4% by 2020, a mere four years away.

In ten years, if we fail to act, ICT will consume about 60 percent of global energy resources. If we manage to optimize energy efficiency in ICT by a factor of 1,000 over 20 years, we’d still see no net change in ICT’s energy use or emissions.

Global use of the Internet is projected to increase 30% to 40% per year. In 10 years, that would mean the Internet will see 30 times its current traffic. That equates to 1,000 times its current traffic in 20 years.

Clearly, the growth in ICT’s energy use and resulting carbon emissions is unsustainable. Because the Internet runs on ICT infrastructure, exponential growth in its use would add another entire dimension to the problem.

Thus when we talk about the advent of the IoT, with billions of devices – the number is projected to rapidly surpass the planet’s current population of 7.3 billion – all talking to each other, these trends accelerate to the point that radical advances in Green ICT are most urgently needed.

Put simply, ICT in the past was designed and built for performance and, henceforth, it needs to be designed for energy efficiency and sustainability.

Question: What specific challenges and opportunities come into play in this scenario?

Cavdar: The biggest challenge, in my view, is that we need 1,000 times more network capacity in the coming 20 years, but with the same energy consumption as today. Many ICT networks will serve consumers at the edge of a network of networks and thus must be affordable for the mobile telecom network operators who build them. They’ll need to profit as they deploy ever- higher capacity networks.

On the opportunity side, ICT is being applied to create more energy efficient processes in all aspects of our lives and across industry verticals. At the same time, ICT itself must become more energy efficient as we deploy more ICT capacity.

Question: Does the pursuit of greener ICT and the application of ICT to make every domain greener, more energy efficient, demand a multidisciplinary effort?

Cavdar: Yes. It requires collaboration and coordination across disciplines, which is a core objective of the IEEE Green ICT Initiative. The Initiative is bringing together a very diverse set of technical domains within and outside of IEEE. Energy efficiency and reduced greenhouse gas emissions are challenges within everyone’s technical domain and that means solutions in certain domains may impact solutions in other domains. The Initiative aims to raise awareness of the challenges outside of everyone’s domain of expertise and to connect individuals in each domain with their colleagues in other domains.

Question: Burak’s work focuses in part on energy efficiency and sustainability of the cloud, whose role would expand considerably as the IoT expands. Will the cloud help or hurt sustainability for ICT and the IoT?

Kantarci: The cloud, like Cicek’s data networks, cuts both ways. The upside is that the cloud offers a single, shared pool of resources to many clients, which in itself is a form of efficiency. And the massive data centers that serve the cloud are, by and large, run in a sustainable manner because energy efficiency and the use of renewable energy is beneficial to the bottom line and cloud vendors understand their environmental responsibilities for an energy-intensive cloud.

Yet accessing the cloud means connecting your network to a data center in a remote location, which increases network traffic in the backbone, creating an energy efficiency challenge. Re-engineering the backbone for greater energy efficiency is a high priority in the Green ICT effort and our thinking includes the use of software-defined networks and network virtualization that can separate the control plane from the data transport plane.

Question: If the edge of the cloud partly involves sensors and devices directly used by humans – in the IoT, the handheld smartphones we use today may give way to wearables or even embedded devices – what are the Green ICT implications of social media?

Kantarci: The fact that most of us carry one or more devices that communicate data offers some opportunities. I currently lead a project sponsored by the National Science Foundation focused on mobile crowd-sensing via social networks. The use of participating individuals’ devices as a mesh or cloud for sending data more efficiently than a single device transmitting to a central network node is a great idea. Yet we still will have to address how to make each of those currently battery-powered, grid-recharged devices more efficient and lower power, perhaps through energy harvesting or other means.

Speaking more generally of the edge of the cloud, I’ve recently become aware of a promising development we refer to as no-power sensors. My colleague, Dr. Silvana Andreescu, in the chemistry and bio-molecular sciences department at Clarkson University, has developed paper-based portable sensors that don’t consume energy but can detect changes in their environment. They can be networked via a mobile backbone. These sensors might replace conventional sensors used in environmental and other types of monitoring.

This is an excellent example of Cicek’s point about the benefits of multidisciplinary, cross-domain efforts.

Question: Cicek, your research has addressed managed versus unmanaged systems and the implications for sustainable networks and Green ICT. Are there implications for the IoT?

Cavdar: Yes. The IoT as we envision it today is becoming a system of systems, a network of networks. The distributed nature of this model makes it more unmanaged by definition – it is controlled by different entities residing in different domains. Those interconnected domains include smart power, smart cities, smart communications, smart transportation, smart manufacturing – you name it. As these domains grow, they consume energy. As we attempt to design a sustainable, largely unmanaged system of systems, we must take a holistic view across these domains and their associated industry verticals because each affects the other, especially in the energy they consume.

Standardization and interoperability is a crucial element in this effort. At this point, multi-system standardization that leads to sustainability and energy efficiency is in its infancy, but we’re aware that this needs to be addressed and initiatives are underway.

Question: What are the short-term and long-term strategies to achieve greener ICT?

Cavdar: As I mentioned earlier, our current ICT was designed for performance rather than efficiency. One short-term strategy is to modify existing systems to make them more efficient, for instance by hibernating or sleeping more when not in demand. Longer term we need to re-design systems to be intelligent, adaptive and optimized for both performance and efficiency.

Kantarci: I agree with Cicek. In the short-term, one specific example I would add is energy-saving schemes in data acquisition and in communications. In the mid-term I would include energy harvesting, which could be vastly improved over what we do now. In the long-term we’ll need no-power, battery-less devices. Such advancements and others could have the transformative impact on Green ICT that we so urgently need.


Read more IEEE Talks IoT articles.