Spectral density constraints on wireless communication

Abstract

Environmental exposure to man-made electromagnetic field (EMF) has been rising as modern technologies have grown and changes in social behavior have generated more synthetic sources. For safety of human health, EMF levels need to be regulated. The level of EMF should be well below levels where there might be harm, hence we do not expect to see any health effects at these levels. Current regulations fail to place a strict limit on EMF in situations where multiple nearby devices transmit simultaneously. The way these regulations are expressed needs great care because it will have an effect on the design of wireless communication systems. In this paper, it is argued that transmitted power constraints on wireless communication devices should be expressed in a different way, namely that devices should limit the EMF spectral density that they generate to the difference between the maximum allowed, by the standard, and the amount currently present, as measured by the device, in the spectral region where it is active. Note that the limit on EMF should be expressed in terms of its EMF spectral density rather than as a total EMF over each of a series of separate bands. If all devices limit their own EMF spectral density, in the spectral region where they are active, in such a way that total EMF spectral density is below the regulated limit in that region, then it is certain that the aggregate EMF spectral density will be below the regulated limit at all frequencies.

Massive MIMO With Radio Stripes for Indoor Wireless Energy Transfer

“…Current regulations impose much more stringent EMF radiation constraints. For instance, ICNIRP establishes the power density safety threshold at ∼ 9.9 W/m 2 (61 V/m)[47], which has been depicted in Fig.9as well. Even so, we adopt Θ 1 = 1000 W/m 2 , above which there might be skin affectations[22], for our discussions in the next subsection since using smaller values would demand operating at higher frequencies, which is extremely computationally demanding and thus requires further and dedicated research.…”

“…As the operation frequency increases, the LOS channel becomes more oscillatory in space, thus, making rather challenging to efficiently perform the numeric integration in ( 38), ( 47) and ( 48), even on high-tech computing servers available nowadays, taking weeks to finalize. In this work, we focused on the 4 GHz operation and thus adopted a relatively high power density threshold, i.e., Θ 1 = 1000 W/m 2 (even when current regulations impose much more stringent constraints, e.g., ∼ 9.9 W/m 2 (61 V/m) [47]) to enable feasibility. As operation at higher frequencies may be a must, future research can focus on addressing above issues, e.g., by using principal component analysis to reduce the dimensionality of the data sets without seriously compromising accuracy [48].…”

Massive MIMO with Radio Stripes for Indoor Wireless Energy Transfer

“…As the operation frequency increases, the LOS channel becomes more oscillatory in space, thus, making rather challenging to efficiently perform the numeric integration in ( 43), ( 52) and (53), even on high-tech computing servers available nowadays, taking weeks to finalize. In this work, we focused on the 4 GHz operation and thus adopted a relatively high power density threshold, i.e., Θ 1 = 1000 W/m 2 (even when current regulations impose much more stringent constraints, e.g., ∼ 9.9 W/m 2 (61 V/m) [43]) to enable feasibility. As operation at higher frequencies may be a must, future research can focus on addressing above issues, e.g., by using principal component analysis to reduce the dimensionality of the data sets without seriously compromising accuracy [44].…”

“…For instance, the radio stripes system may need just 25 dB to serve 8 Current regulations impose much more stringent EMF radiation constraints. For instance, ICNIRP establishes the power density safety threshold at ∼ 9.9 W/m 2 (61 V/m) [43], which has been depicted in Fig. 8 as well.…”

“…which comes from using Hk = r k 0 H∫ k (r)dr, where H∫ k (r) is defined after (43), and from using P (x, y, z) given in (47) and setting H =…”