“There’s no doubt that benefits of IoT are multi-dimensional. However, specific solutions must be designed efficiently, to become self-sustainable”

 
Onel López

In this interview, Onel López, Assistant Professor in Sustainable Wireless Systems at the University of Oulu, discusses his work on sustainable IoT systems, the challenges arising from massive, machine type communications and the innovation opportunities arising from 5G and 6G.

Q1: Would you begin with some introductory information about yourself and how you come to be based in Finland?

OL: I am originally from Cuba where I worked as a telematics specialist for ETECSA, the Cuban telecommunication company, after completing my studies in Telecommunication and Electrical Engineering at the Central University of Las Villas. I then moved to Brazil, to complete my master’s studies at the Federal University of Paraná. That is where I started my research on low-power communications, specifically on RF wireless power transfer for powering IoT deployments.

With the help of my supervisors Prof. Evelio Fernández and Prof. Hirley Alves, and collaborators Dr. Samuel Montejo-Sánchez, Dr. Samuel Mafra, Prof. Richard Souza, and Prof. Glauber Brante, I published early results of my work in prestigious IEEE journals and conferences. That led to an offer from Prof. Hirley Alves and Prof. Matti Latva-aho to carry out my doctoral studies and research at University of Oulu (UOULU), Finland. My PhD studies and research were focused on resource allocation for machine-type communications, which are at the core of current and future IoT deployments. Things progressed well and after 3 years I defended my doctoral thesis with honors. I also received a one-of-a kind offer to join UOULU tenure’s track as Assistant Professor in Sustainable Wireless Systems.

I must admit that I never pictured myself in a country like Finland. It is quite the opposite in terms of nature, weather, and social environment compared to Cuba and Brazil. Finland has been a much more welcoming place than I anticipated. I really appreciate every aspect of life here, especially the world-class research machinery, infrastructure, social and nature harmony, and well-being.

Q2: You won an award for the best doctoral thesis in 2021 on “Resource allocation for machine-type communication”.  What were the main issues you set out to explore?

OL: We are seeing a rapid proliferation of the IoT and associated use cases, which is leading to an astonishing increase in the number of wirelessly connected devices. There is a high degree of heterogeneity, both in terms of characteristics and requirements. These developments pose significant challenges to current machine-type connectivity solutions. To understand this better, it is useful to think about two main machine-type communication (MTC) regimes. One is critical MTC, which is concerned with providing ultra-reliable low-latency communication (URLLC) in controlled environments with small-payloads and low-data rates. The second is massive MTC, which aims to serve large and dense IoT deployments with sporadic traffic patterns and to do so in ways that use energy efficiently.

There are huge challenges in addressing such MTC scenarios for next generation of wireless networks. It requires new approaches, methods, and ideas. Keep in mind that most MTC devices use simple hardware and are energy limited, while wireless networks are much more complex to handle multiple heterogeneous services. There is therefore a fundamental need for efficient resource allocation strategies to handle stringent latency, reliability, energy efficiency, and/or service availability constraints.

In my doctoral thesis, I explored several resource allocation strategies including power/rate control, relaying cooperation, spatial diversity, aggregation, and non-orthogonal multiple access. I placed special emphasis on improving energy efficiency. This is crucial for IoT since it prolongs the lifetime of networks and devices, reduces the need for human intervention, lowers maintenance costs and improves the user experience. In this regard, we introduced the concept of massive wireless energy transfer, which refers to wirelessly charging massive low-power IoT nodes’ deployments via RF signals.

Q3: What factors will drive demand for the new IoT scenarios and what challenges will these raise?

OL: Last year, I worked with a group of experts in networks, business, wireless communications, and other related fields, to shed light on this very question. This formed part of the 6G White Papers published by the University of Oulu. In our view, the factors are many and not all of them fully identified or understood. Nevertheless, here are some of the most representative.

There is a trend to turn our homes, entertainment, health, work, social/community services into smart and interactive environments. This requires large, and preferably imperceptible, IoT deployments that support ubiquitous connectivity anywhere and anytime. Augmented/ virtual/ mixed reality (XR) may be key for this. MTC will play a fundamental role in facilitating the user experience and enabling novel and efficient human-machine interfaces to present the data coming from machines in a more natural way. Rapid advances in wireless brain-computer interactions and multisensory XR requirements will drive a massive proliferation of cost-effective miniaturized smart wearable and implant sensors.

Moreover, increasing the number of sensors and edge processing systems are being integrated within vehicles. The same applies to other mobile nodes and this is paramount for automation and autonomous mobility. Given the proliferation of unmanned vehicles and related applications, we also need to reconsider the connectivity landscape. Something that was once bi-dimensional and is now also focused on heights.

In industry and factory settings, IoT will facilitate the production of customized and personalized products in mixed sensing/actuation/haptics scenarios. This is quite different from the dominant ‘data collection and analytics’ scenarios of nowadays. Interestingly, the data generated by widely distributed IoT nodes will have enormous business, environmental, and societal values that can be intelligently leveraged to promote inclusiveness, trustworthiness and (self-)sustainability.

I commented earlier that IoT connectivity is nowadays focused on two main regimes: critical MTC and massive MTC. Over the coming years, however, this distinction will blur because of emerging use cases and the verticalization of service provision. These will demand stricter and higher-level service guarantees, including multi-dimensional optimization and scalable design. One of the key challenges ahead is to address the coexistence of heterogenous IoT services, with potentially conflicting requirements, while delivering uninterrupted as well as cost and energy-efficient operational guarantees on a massive scale.

Q4: Speaking about a massive increase in the number of IoT devices, can you illustrate what that might look like?

OL: The massive growth projections for the IoT are indeed astonishing. Take for instance https://www.ericsson.com/en/mobility-report/mobility-visualizer the more conservative. According to this, the number of wide-area and short-range IoT devices will double within the next 4 years! If we set our sights a bit further towards 2030 and the 6G era, the numbers may be quadrupled or sextupled.

The goal set by academia and industry with regards to efficiently supporting the operation and connectivity of 10 IoT devices per square or 100 per cubic meter by 2030 is not exaggerated at all. In an area of 10-meter radius, this would translate to supporting the coexistence of more than 3000 IoT devices!

These projections raise two concerns. One is the massive access problem, and the other relates to constrained IoT devices and their service-life consequences. Industry and academia have started to investigate solutions based on modern random-access protocols, fast uplink grants, intelligent grant-free mechanisms, sensor fusion and aggregation techniques.

Since traditional sub-6GHz spectrum is scarce, exploiting millimeter-wave, and even THz, frequency regions may be compelling. However, complexity and energy consumption scale with the operating frequency, which poses serious obstacles here.

Finally, the deployment of (traditionally) battery powered IoT devices does not scale well. It might also not be sustainable, considering the frequency of battery or device replacements and the contribution to e-waste. To tackle this issue, energy-harvesting from green energy sources and wireless power transfer have emerged as appealing solutions.

Q5: What is involved in delivering reliable connectivity? Is there a balance of technical and operational issues and do these vary in non-time-critical and low-latency situations?

OL: Critical MTC/IoT use cases, as in industrial automation and vehicular communications, demand both ultra-high reliability and low latency guarantees. If one considers an infinite time horizon, delivering reliable connectivity is not really an issue. Different forms of diversity, i.e., time, frequency, spatial, can be leveraged as network and spectrum resources become available over time. The problem becomes relevant, and highly cumbersome, when a critical time/latency constraint is imposed since achievable reliability decreases as available communication time shrinks.

When it comes to providing strong reliability guarantees, there is a need for accurate statistical characterization of critical phenomena such as interference, operational failure, and other rare events. Unfortunately, this is not always possible, especially in dynamic scenarios, where such statistics are time varying. Therefore, although such statistics may be (accurately) estimated, the reliability guarantees they provide must reveal the underlying assumptions and uncertainties transparently.

Operating in a low-latency regime imposes critical challenges of its own. Specifically, traditional communication protocols are designed for long-codeworks, where message metadata is considerably smaller than the payload. This may not hold at all in critical MTC. In other words, adopting a similar message structure in short packet communications leads to highly suboptimal reliability performance. Note however that such metadata or overhead is nevertheless useful for channel state information acquisition, to identify and authenticate devices, and for other signaling purposes. One cannot completely avoid it so joint metadata-payload encoding may still be needed.

In a similar fashion, the traditional four-message handshaking for random access (before establishing a functional communication link) is unaffordable under tight latency constraints. Fixed resource scheduling may be inefficient given sporadic IoT activation patterns, as in IoT deployed for detecting rare events to trigger alarms. These are a few examples to illustrate why there is research into intelligent fast-uplink and resilient grant-free mechanisms. These issues demand a complete re-thinking of wireless communication protocols to support critical MTC.

Finally, although our discussions were mostly focused on the lower communication layers, the problem scales up when moving to upper communication layers and offering end-to-end performance guarantees.

Q6: With growing interest in sustainability via IoT, would you tell us a bit about your work on green communications and promising areas of research?

OL: IoT can indeed be leveraged to promote sustainability, and you often hear the term “sustainability via IoT”. There are many examples where IoT sensors make a difference with examples such as detecting people leaving a room to turn off the lights and save energy, or to identify air/water contamination levels and call for intervention. There is no doubt that the potential and benefits of IoT are multi-dimensional. However, IoT specific solutions must be designed efficiently to become self-sustainable, and the main motivation of our work is precisely this: supporting sustainable IoT.

The main work my team and I are carrying out relates to sustainable IoT enabled by energy harvesting and wireless energy transfer. In fact, we recently compiled a book on the topic. I should point out that energy harvesting techniques are attractive for recharging batteries wirelessly and to avoid replacement. In addition to being costly, replacement might be impossible in hazardous environments, in building structures or for sensors in the human body. Among the range of energy sources, we consider RF signals which can be harvested simultaneously by several devices.

Since RF energy availability and density fluctuate in space and time, and may be strictly limited, we see a need to deploy dedicated energy transmitters that can charge IoT devices to support quality-of-service goals. However, this comes at the expenses of introducing new devices into the network, with their associated carbon footprint, communication functionalities and protocol overhead.

Another less popular but equally relevant aspect of sustainability lies in controlling RF pollution. RF transmission in microwave and mm-wave bands are non-ionizing and thus unable to change the structure of atoms and molecules with potentially carcinogenic effects, for example. Nevertheless, several safety limits have been set by international regulatory bodies to limit the electromagnetic field (EMF) exposure and to avoid biological effects such as tissue heating. This, together with the growing concerns of people over exposure to radio waves, which are rising with the advent of the 5G wireless systems and the exploitation of higher frequency bands, call for transparent mechanisms to ensure compliance with EMF exposure regulations.

All these aspects, including mechanisms for boosting the end-to-end power transfer efficiency, demand a profound study, analysis, and holistic optimization approaches to promote system sustainability.

Q7: Among the different applications and technologies you have covered, in which areas do you see opportunities for standardization?

OL: In general, standardization of MTC is expanding, and will continue to do so because of the massive penetration of IoT technologies and emerging use cases with increasingly stringent performance requirements. I see special potential for standardization in the areas of: i) random access protocols for massive and critical MTC, ii) new IoT device classes (e.g., RF energy transmitters, intelligent reflective surfaces, backscatter nodes, THz nodes, satellites), iii) indoor/outdoor accurate localization and sensing, iv) zero-energy computing/communication technologies, v) security, authentication and resilience mechanisms, vi) non-terrestrial connectivity, vii) edge computing and intelligence, viii) energy harvesting and trading, ix) collaborative network slicing, x) massive MIMO for MTC, xi) non-RF communication technologies, and xii) multi-RAT operation and coexistence. This is by no means an exhaustive list.

Although several small, medium, and large-scale enterprises such as TransferFi, PowerCast, Energous, Humavox, Ossia, Guru, Xiaomi, have taken a leap forward in the RF wireless energy transfer technology development, standardization in the area is in fact very incipient. The main standardization attempt occurred in 2015, when Alliance for Wireless Power (A4WP) and Power Matters Alliance (PMA) merged with the goal of facilitating the collaboration between the different technologies, so that end-users will not even know which technology they are using to charge wirelessly. However, such an effort has not taken off yet, possibly because of technology limitations. Looking ahead, advances in technology and standardization would both speed up the path to mature technologies while building trust in EMF exposure compliance.

Q8: As you are part of the 6G Flagship Project at the University of Oulu, what ideas excite you about 6G in general and its importance for IoT?

OL: There is a huge focus, by industry and academia, on the idea of a data-driven sustainable future society, enabled by near-instant, secure, unlimited, and green connectivity. Since IoT is a fundamental enabler of such a vision and a lucrative and expanding business area, the cellular industry and proprietary solutions started developing efficient support for it. Specifically, 5G, with an incipient vertical-oriented service provisioning, expanded the traditional human-type connectivity scope of the cellular industry by including support for a variety of challenging MTC/IoT use cases. In parallel, SigFox, LoRA/LoRAWAN and Ingenu led the low-power wide area network solutions.

As the technologies that support IoT systems continue to evolve, there is a need for more robust, scalable, and efficient infrastructure designs to address current limitations. A holistic 6G infrastructure will indeed offer unprecedented, native support for IoT, while meeting the emerging requirements of 2030 society and beyond. The goal, the challenges, and the whole road ahead, are all exciting for somebody like me who enjoys mathematics, research, technology, nature, and life. Moreover, being part of the 6G Flagship research community at the University of Oulu is a marvelous experience in many areas, both professional and personal. We are hundreds of passionate people trying to shape/enable, prepare for, and look objectively into the future of wireless communications.