Petar Popovski, Editor in Chief of IEEE JSAC and part of the Guest Editorial Team
Published: 6 Mar 2023
As the deadline for submission to the special issue on Electromagnetic Signal and Information Theory is approaching, we have reached out to the Guest Editors to ask for a perspective on this emerging research area by asking three questions, some of them rather provocative. The responses were provided by a part of the Guest Editorial Team: Rodrigo C. de Lamare (RCdL), Kumar Vijay Mishra (KVM), Michalis Matthaiou (MM), and Gerhard Kramer (GK). We believe the readership will find these responses to be insightful and stimulating for further exploration of the research problems in this area.
JSAC: Information theory is quite a general mathematical framework. For instance, in information theory it is known that "channel is a part of the system we are unwilling or unable to change", but once the channel is defined, we can use Shannon theory to analyze its capacity and error probability. Why, then, do we need information theory tailored to the electromagnetic features of the signals?
RCdL: Taking into account the electromagnetic features of the signals has the potential to bring more accurate information theoretic results and eventually improve the design of communication systems.
KVM: One of the major drivers behind the recent interest in electromagnetic signal and information theory (ESIT) is the ongoing active research in metasurfaces. These devices comprise an array of spatially-varying sub-wavelength scattering elements that can control and manipulate electromagnetic waves through modified surface boundary conditions. Hence, it becomes pertinent to study the information theory of metasurface-based wireless communications through the lens of electromagnetics. These ESIT insights then naturally extend to several other applications where the system performance is directly linked with electromagnetics of the devices or channel, such as near-field architectures, terahertz electromagnetics, and even radio wave interaction with biological systems.
MM: While information theory is undoubtedly a versatile mathematical tool, it is based on mathematical logic. This theoretical framework now needs to be extended and reshaped to incorporate the main feature of future communication systems, namely their capability of sensing the system’s response to the radio waves, and thereby informing its modification.
GK: The statement suggests that information theory is limited to “channels”, “capacity”, and “error probability” but this is a narrow perspective. Information theory addresses the design of the entire communications chain from the application layer (video coding, speech coding, bit coding, security coding) to the physical layer (FEC, feedback, multi-carrier signaling, power control) for all block lengths. Moreover, multi-user information theory provides the right framework for multi-user systems (multi-access, random access, spatial multiplexing). For instance, the optimality of separating source and channel coding for point-to-point links suggested a layered approach to communications that continues to be important today to ensure flexibility. And the refined statement that one cannot separate source and channel coding for multi-user systems motivates cross-layer improvements. Inevitably, we must combine information theory with channel modeling. This happened for telephone channels (1980s-2000s), wireless channels (1990s-2010s), and fiber channels (2000s-2020s). The further development of this fascinating synergy is one goals of this issue.
JSAC: Can you give an example of a research question in Electromagnetic Signal and Information theory that you see as particularly challenging or interesting?
RCdL: An interesting challenge of Electromagnetic Signal and Information theory is to come up with sufficiently simple models that accurately capture impairments and practical effects related to Electromagnetics, which would then allow the development of cost-effective signal processing solutions and yet facilitate the use of information theory to provide better insights on system design than those we currently have.
KVM: Yes. A particularly interesting research problem could be to include the electromagnetic parameters such as the operating frequency, polarization, dielectric properties, and permeability of the metasurfaces directly into the information-theoretic analysis of metasurface-aided wireless communications.
MM: As the number of antenna elements in future arrays increases (e.g XL MIMO, THz arrays), their near-field zone will continue to expand. To date, near-field communication has been a black box for information theorists. In order to understand the impact of near-field propagation (e.g. non-stanionarities, spherical wavefronts), we need to go back to our information theoretic models and recalibrate them using these unique electromagnetic characteristics.
GK: A simple and obvious answer is: all practical channels, including copper, optical fiber, wireless, magnetic, free-space optical, etc. Combining channel modeling and information theory is particularly challenging because both theories are sophisticated on their own.
JSAC: During the past 25 years we have seen a continuous evolution in the area of multi-antenna communication and witnessed the direct impact of those techniques to practical systems and standards. Do you see a potential in Electromagnetic Signal and Information theory, over a longer run, to trace an evolution path that will lead to new waveforms, new types of devices or new types of network elements? Can you elaborate upon that?
RCdL: There is tremendous potential in Electromagnetic Signal and Information (ESIT) theory to better inform designers about how to obtain performance benefits by using more realistic modeling of multiple-antenna systems and networks. In particular, when a designer employs ESIT theory on a particular network element such as an access point or an intelligent reflective surface it could potentially provide more clear directions on how to obtain performance benefits.
KVM: Yes. ESIT has the potential to enable more accurate and reliable characterization of communications systems. At present, the multi-antenna multi-functional metasurface-aided systems are already being evaluated by start-ups and even some large telecommunications organizations. Future wireless communications are likely to employ a suite of metasurface-aided techniques (e.g., near-field, THz, OAM, and holographic systems) that must also have an electromagnetically-compliant design.
MM: Multiple-antenna communications have indeed witnessed an unprecedented evolution over the past 25 years. The emerging multiple-antenna technologies, such as THz communications, RIS and OAM, create new challenges in their modeling, transceiver design/optimisation and waveform design. In this context, injecting electromagnetic knowledge into the conventional Gaussian-type information theoretic models becomes a timely and extremely interesting exercise.
GK: Yes, only foundational theories can support sustainable progress in communications. History has taught us that electromagnetic theory and information theory are two of those theories, and that they complement each other.
Statements and opinions given in a work published by the IEEE or the IEEE Communications Society are the expressions of the author(s). Responsibility for the content of published articles rests upon the authors(s), not IEEE nor the IEEE Communications Society.