TECHNICAL FIELD

Various examples of the disclosure are broadly concerned with coverage enhancing devices (CEDs).

BACKGROUND

In order to increase a coverage area for wireless communication, it is envisioned to use coverage enhancing devices (CEDs), particularly reconfigurable relaying devices (RRD), more particularly, reconfigurable reflective devices. Reconfigurable reflective devices are sometimes also referred to as reflecting large intelligent surfaces (LISs). See, e.g., Sha Hu, Fredrik Rusek, and Ove Edfors. "Beyond massive MIMO: The potential of data transmission with large intelligent surfaces." IEEE Transactions on Signal Processing 66.10 (2018): 2746-2758.

An RRD can be implemented by an array of antennas that can reflect incident electromagnetic waves/signals. The array of antennas can be semi-passive. Semi-pas- sive can correspond to a scenario in which the antennas can impose a variable phase shift and typically provide no signal amplification. An input spatial direction from which incident signals on a radio channel are accepted and an output spatial direction into which the incident signals are reflected can be reconfigured by changing a phase relationship between the antennas. Radio channel may refer to a radio channel specified by the 3GPP standard. In particular, the radio channel may refer to a physical radio channel. The radio channel may offer several time/frequency-re- sources for communication between different communication nodes of a communication system.

Sometimes CEDs may also be referred to as Large Intelligent Surfaces (LIS) and Re- configurable Intelligent Surfaces (RIS). CEDs are foreseen to be a part of 5G+ and 6G systems to be specified by the 3GPP. They have spurred significant research efforts from both academia and industry. Using CEDs may in particular be a low cost, low power approach to improve coverage, capacity, and avoid shadowing. Most research efforts target reflective CEDs but transmissive CEDs (or lens-arrays) are also being investigated. Especially for systems operated at mm-wave frequencies, where dispersion is significantly lower compared to sub 6 GHz bands, the shadowing is troublesome. The CEDs are in some contexts seen as a new type of node next to access points and terminals in the networks. CEDs may be similar to amplify & forward repeaters in its function, but may need additional control signaling for the beam configuration.

Both reflective and transmissive surfaces can be comprised of what is known as meta-surfaces or meta-material. In general, a meta-material is a material that has a refraction- or reflection-index which changes the reflected propagation direction of an impinging electro-magnetic wave compared to Fresnel’s law. Typically, a meta-mate- rial is designed to either reflect or act transmissively to an impinging wave. The metasurface is additionally designed to be configurable. By configuring the properties of the surface the index can be changed and an arbitrary reflection or refraction angle of departure (AoD) can be configured.

SUMMARY

There may be a need for CEDs providing more coverage than known CEDs.

This need is met by the features of the independent claim. The features of the dependent claims define examples.

Examples provide a coverage enhancing device (CED) comprising a first panel being reconfigurable for changing an angle of refraction of incident signals received on a radio channel and transmitted through the first panel and a second panel being re- configurable for transmitting the incident signals received via the first panel through the second panel or reflecting the incident signals received via the first panel.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic side view of an MFSS according to various examples.

FIG. 2 schematically illustrates an array of metallic elements of a top layer and a bottom layer of the MFSS according to various examples.

FIG. 3 schematically illustrates an array of metallic elements of a top layer and a bottom layer of the MFSS according to various examples. FIG. 4 schematically illustrates an array of metallic elements of a middle layer of the MFSS according to various examples.

FIG. 5 schematically illustrates an array of metallic elements of a middle layer of the MFSS according to various examples.

FIG. 6 schematically illustrates an array of metallic elements of a middle layer of the MFSS according to various examples.

FIG. 7 schematically illustrates an array of metallic elements of a middle layer of the MFSS according to various examples.

FIG. 8 schematically illustrates an array of metallic elements of a middle layer of the MFSS according to various examples.

FIG. 9 schematically illustrates an array of metallic elements of a middle layer of the MFSS according to various examples.

FIG. 10 is a frequency response of the MFSS including a passband and a stopband according to various examples.

FIG. 11 is a frequency response of the MFSS including a passband and a stopband according to various examples.

FIG. 12 is an equivalent circuit model of the MFSS according to various examples.

FIG. 13 schematically illustrates a tunable capacitor arranged in the gap between metallic elements of a middle layer of the MFSS according to various examples.

FIG. 14 schematically illustrates biasing the tunable capacitors according to various examples.

FIG. 15 schematically illustrates a system including the MFSS attached to a cover of a housing of a wireless communication device.

FIG. 16 is a flowchart of a method according to various examples.

FIG. 17 schematically illustrates a CED.

FIG. 18 schematically illustrates a CED. FIG. 19 schematically illustrates a CED.

FIG. 20 schematically illustrates a CED.

FIG. 21 schematically illustrates a CED.

FIG: 22 schematically illustrates a CED.

FIG. 23 schematically illustrates a scenario for using a CED.

DETAILED DESCRIPTION

Frequency-selective surfaces

Glass is a widely used material for housings of mobile phones. However, glass can have a significant impact on the propagation properties of electromagnetic waves, in particular in the millimeter wavelength regime. This is because glass has a high permittivity. For instance, the permittivity (relative to the permittivity of vacuum) of glass can be between 5.5 and 7 which can be twice as large if compared to plastic. Plastic can typically have a permittivity below 3.0.

Thus, glass can significantly block electromagnetic waves used for communication purposes. For instance, it has been observed that for electromagnetic waves having a frequency of 28 GHz, transmission can be completely blocked at certain incident angles.

To improve the transmissivity, frequency-selective surfaces (FSSs) can be used. FSSs are disclosed in:

Anwar, Rana Sadaf, Lingfeng Mao, and Huansheng Ning. "FSSs: a review." Applied Sciences 8.9 (2018): 1689.

Chang, Kihun, Sang il Kwak, and Young Joong Yoon. "Equivalent circuit modeling of active FSSs." 2008 IEEE Radio and Wireless Symposium. IEEE, 2008.

Bayatpur, Farhad, and Kamal Sarabandi. "Tuning performance of metamaterialbased FSSs." IEEE Transactions on Antennas and Propagation 57.2 (2009): 590- 592. Al-Joumayly, Mudar, and Nader Behdad. "A new technique for design of low-profile, second-order, bandpass FSSs." IEEE Transactions on Antennas and Propagation 57.2 (2009): 452-459.

Hereinafter, techniques associated with FSSs will be disclosed. An FSS is a passive or semi-passive device that exhibits a(adjustable) response to electro-magnetic radiation. In particular, the FSS may exhibit an adjustable frequency response, i.e. a response depending on the frequency of an electro-magnetic wave. To tailor frequency-selective transmission/reflection characteristics of the FSS, a planar and periodic array of metallic elements may be used. A thickness of a corresponding layer may be negligible when compared to the wavelength. However, the thickness of the layer can be larger than the skin depth of the metal. Thus, such array of metallic elements can be approximated as an array of perfectly conducting resonant elements. A periodicity of the metallic elements in the respective array may be on the same order of magnitude as the wavelength, or even smaller.

According to various examples, a multi-layer frequency-selective surface (MFSS) is used. Here, a bottom layer, a top layer and a middle layer are arranged in between the top layer and the bottom layer.

According to various examples, a tunable MFSS is disclosed. Here, tunable capacitors can be used to adjust a frequency response of the MFSS. In particular, the frequency of a stopband of the MFSS can be adjusted by applying an appropriate bias voltage to the tunable capacitors. For example, voltage-controlled capacitors could be used or a PIN diode.

An MFSS according to the disclosed examples can include a stopband in its frequency response. The stopband can help to reduce interference. Out-of-band emissions can be reduced which limits exposure of adjacent biological matter to electromagnetic waves.

According to various examples, MFSSs disclosed herein can have a tailored frequency of the stopband. Firstly, a degree of freedom of design parameters of the geometrical structures used to implement the MFSS can be sufficiently large so as to enable flexible adjustment of the frequency of the stopband without impact on the frequency of an adjacent passband. For instance, the frequency of the stopband may be tuned by the choice of one or more of the following design parameters of the array of metallic elements of the middle layer: shape of the metallic elements; width to periodicity ratio of the metallic elements; gaps between the metallic elements. Change of such parameters may not or only negligibly affect the frequency of the passband. Furthermore, for tunable FSSs, it is possible to dynamically tune the frequency of the stopband without affecting the frequency of the passband.

According to techniques disclosed herein, it is possible to implement multiple stopbands at different frequencies for horizontal and vertical polarizations of the electromagnetic waves. For instance, a twofold rotational symmetry of the array of metallic elements of the middle layer may result in different stopbands for the two polarizations. The twofold symmetry can be achieved by using, e.g., different gap widths for gaps along X-direction and Y-direction of that array.

FIG. 1 schematically illustrates an MFSS 110. The MFSS 110 includes a top layer 141 , a middle layer 142, and a bottom layer 143 (“top” and “bottom” are arbitrarily defined). The top layer 141 includes an array 151 of metallic elements. The middle layer 142 includes an array 152 of metallic elements. The bottom layer 143 includes an array 153 of metallic elements. The arrays 151 , 152 are separated by a dielectric layer 161. The arrays 152, 153 are separated by a dielectric layer 162.

Illustrated in FIG. 1 are incident electromagnetic waves 121 , reflected electromagnetic waves 122, as well as transmitted electromagnetic waves 123. A ratio of the energy or amplitude of the transmitted electromagnetic waves 123 to the energy or amplitude of the incident electromagnetic waves 121 defines a transmissivity of the MFSS. A ratio of the amplitude or energy of the reflected electromagnetic waves 122 to the energy or amplitude of the incident electromagnetic waves 121 defines a reflectivity of the MFSS. For example, the array elements may have a size in the order of half of the wavelength of incident electromagnetic waves. This would render it possible to let electromagnetic waves pass through or being reflected the structure by switching the array elements on or off.

FIG. 2 schematically illustrates an example implementation of the arrays 151 , 153 of the top layer 141 and the bottom layer 143. For instance, it would be possible that the array 151 is configured the same as the array 153. In the example of FIG. 2, the arrays 151 , 153 are implemented by capacitive patches as metallic elements 231 , separated by gaps along an x-direction and a y-direction.

The arrays 151 , 153 have fourfold rotational symmetry in the illustrated example. Thus, a passband position for horizontally and vertically polarized electromagnetic waves would be the same.

It would also be possible to use different gap widths in the x-direction and/or the y- direction, e.g., resulting in twofold rotational symmetry. Then different passbands can be configured for horizontally and vertically polarized electromagnetic waves.

FIG. 3 schematically illustrates an example implementation of the arrays 151 , 153 of the top layer 141 and the bottom layer 143. The arrays 151 , 153 are implemented by capacitive circles as metallic elements 232 (forming the unit cell of the arrays 151 , 153). The arrays 151 , 153 again have a fourfold rotational symmetry, which is generally optional.

As will be appreciated, in FIG. 2 and FIG. 3, the metallic filling fraction of the arrays 151 , 153 is comparably high, e.g., higher than 95% for FIG. 2 and higher than 70% for FIG. 3. A higher metallic filling factor may provide better efficiency for higher bandwidths. Moreover, a higher metallic filling factor may reduce the variation of the performance for different frequencies. A higher metallic filling factor may lead to arrays 151 , 153 being less dispersive.

FIG. 4 schematically illustrates an example implementation of the array 152 of the middle layer 142. In the example of FIG. 4, the array 152 is implemented by square loop-shaped metallic elements 241 separated by gaps 261 , 262 along both x-direc- tion, as well as y-direction (thereby forming the unit cell of the array 152).

The metal filling fraction of the array 152 is less than 15%. This is, in particular, significantly smaller than the metal filling fraction of the arrays 151 , 153 of the top layer 141 and the bottom layer 143 discussed in connection with FIG. 2 and FIG. 3 above.

Such difference in the metal filling fraction enables accurate modelling of the frequency response of the MFSS using an equivalent circuit model in the form of a Pi- filter network. FIG. 5 schematically illustrates an example implementation of the array 152 of the middle layer 142. In the example of FIG. 5, the array 152 is implemented using rectangular loop-shaped elements (forming the unit cell of the array 152).

These elements are only distanced I separated by the gaps 261 along the x-direction - and joined together along the y-direction.

Accordingly, a two-fold rotational symmetry of the array 152 is implemented.

Accordingly, horizontally (along the x-direction) polarized electromagnetic waves will be affected differently compared to vertically (along the y-direction) polarized electromagnetic waves. For example, one of the two polarizations can experience a significant stopband, where the stopband may not be present or may not be pronounced in the frequency response for the other polarization.

FIG. 6 schematically illustrates an example implementation of the array 152 of the middle layer 142. In the example of FIG. 6, the array 152 is implemented using rectangular loop-shaped metallic elements 242 (forming the unit cell of the array 152). These elements are separated by the gaps 261 along x-direction and separated by the gaps 262 along the y-direction.

The gaps 261 , 262 have different widths (the gaps 261 are wider), which is different to the scenario of FIG. 4.

Accordingly, the frequency of stopband for the horizontally polarized electromagnetic waves is different to the frequency of the stopband for the vertically polarized electromagnetic waves.

FIG. 7 schematically illustrates an example implementation of the array 152 of the middle layer 142. In the example of FIG. 7, the array 152 is implemented using crossshaped metallic elements 243 (forming the unit cell of the array 152). Again, different widths of the gaps 261 , 262 are used, which is generally optional. It would also be possible that the crosses are joined together along the y-direction (not shown).

FIG. 8 schematically illustrates an example implementation of the array 152 of the middle layer 142. In the example of FIG. 8, the array 152 is implemented using Jerusalem crosses as metallic elements 244 (forming the unit cell of the array 152). It would be possible that gaps 261 , 262 of different widths are used. The crosses could be joined together along the y-direction.

The examples as illustrated above implement a two-fold (FIG. 5, FIG. 6, FIG. 7) or four-fold (FIG. 4, FIG. 8) rotational symmetry for the array 152. Also, a higher-order rotational symmetry would be possible, as illustrated in connection with FIG. 9 (where a six-fold rotational symmetry is shown).

FIG. 9 schematically illustrates an example implementation of the array 152 of the middle layer 142. In the example of FIG. 9, the array 152 is implemented using a hexagonal unit cell, sometimes referred to as “3-legged loaded”. Again, cross-shaped metallic elements 245 are used, having three legs. A unit cell results that includes multiple such three-legged crosses.

FIG. 10 illustrates a frequency response 400 of an MFSS 110 according to various examples. A transmission gain is plotted as a function of frequency. The transmission gain corresponds to the transmissivity. Illustrated is a passband 421 and two stopbands 431 , 432 for horizontal polarization and vertical polarization, respectively.

While in FIG. 10 the stopbands 431 , 432 are arranged above the passband 421 , as a general rule, it would be possible that the one or stopbands are also arranged below the passband 421 , as illustrated in FIG. 11.

FIG. 12 illustrates an equivalent circuit model 800 of the MFSS. A Pi-filter passive network is implemented. As illustrated, the top layer 141 and the bottom layer 143 are modeled using parallel capacitances (accordingly, the metallic elements 231 , 232 can be labeled capacitive metallic elements). The middle layer 142 is modeled by an inductance 801 , as well as a capacitance 802 (accordingly, the metallic elements 241-245 can be labeled capacitive metallic elements).

It has been observed that the capacitance 802 is mainly affected by the width of the gaps 261 , 262 in-between adjacent metallic elements 241-245 of the middle layer 142. Also, the capacitance 802 affects the frequency of the respective stopband 431 , 432. Thus, by tuning the width of the gaps 261 , 262, it is possible to tune the frequency of the respective stopband 431 , 432. To be able to tailor the frequency of the stopbands 431 , 432 even more flexibly, it would be possible to arrange a tunable capacitor 701 in one or more of the gaps 261 , 262, as illustrated in FIG. 13. For instance, the tunable capacitor 701 could be implemented as a PIN diode or a voltage-controlled capacitor.

FIG. 14 schematically illustrates a voltage source 710 applying a bias voltage to a series connection of the tunable capacitors 701 (varactors) arranged in adjacent gaps 261 , 262 of the array 152. This simplifies the electrical supply network. While in FIG. 14 the tunable capacitors 701 are illustrated as PIN diodes, other implementations would be similarly possible, e.g., using varactors.

FIG. 15 schematically illustrates a system 900. The system 900 includes a housing 980 of a wireless communication device 981 , e.g., a tablet or a mobile phone. A glass cover 920 is used. Generally, a cover made from a high permittivity material having a permittivity of not less than 4 or 5 could be used. An MFSS 110 is attached to the glass cover 920 in an area adjacent to an antenna 910. A frequency of one or more passbands of the MFSS is matched to a frequency of the antenna 910.

A control unit 985 is provided that can control a voltage source 710 to apply a voltage to the tunable capacitors 701 (not illustrated in FIG. 15). For instance, this can be based on control data that is indicative of a frequency of a desired stopband.

FIG. 16 is a flowchart of a method according to various examples. For instance, the method of FIG. 16 could be executed by the control unit 985 or another processor. A voltage source coupled with tunable capacitors arranged in gaps 261 , 262 as discussed above can be controlled.

At box 3005, control data is obtained that is indicative of a frequency of a stopband. The control data could be loaded from a memory.

At box 3010 a voltage source can be controlled to apply a voltage to the tunable capacitors that are arranged in gaps between adjacent metallic elements of a respective array of a MFSS, e.g., as discussed in connection with the FIGs. above.

Full-sphere CEDs

Fig. 17 shows a CED 1700 comprising a first panel 1710 and a second panel 1720 implemented as a single, combined structure. The first panel 1710 is a refractive meta surface. The second panel 1720 is reconfigurable for transmitting or reflecting an incident signal received via the first panel 1710.

The first panel 1710 comprises conductive elements 1712 and a reflective reconfigurable ground 1711. The phase response of the first panel 1710, which may also be considered as a phase shifting layer, may determine the relation between the angle of arrival (AoA) of an incident signal 1771 and the angle of departure (AoD) of a reflected or transmitted incident signal. The phase response may be controlled by variable capacitors interconnected between the conductive elements 1712 and ground 1711. The phase response may also be controlled mechanically, e.g. by changing the relative position of the conductive elements 1712 and ground 1711.

The second panel 1720 may comprise further conductive elements 1722. The conductive elements 1722 may be controlled such that the incident signal 1771 is either transmitted through ground 1711 or reflected. The second panel 1720 may be implemented in the form of an MFSS as described hereinbefore.

In particular, the CED 1700 may be configured to have different passbands/stop- bands for transmitting and reflecting incident signals.

Fig. 18 illustrates a CED 1800 similar to the CED 1700 comprising a first panel 1810 and a second panel 1820. In contrast to the CED 1700, the first panel 1810 and the second panel 1820 are implemented as separate structures. As explained before, the first panel 1810 may comprise the conductive elements 1812 and ground 1811 and the second panel 1820 may include further conductive elements 1822.

In some examples, the second panel of a CED described herein may be implemented as a tunable FSS as described hereinbefore.

Fig. 19 shows a CED 1900 comprising a first panel 1910 and a second panel 1920. The first panel 1910 is reconfigurable for changing an angle of refraction of incident signals 1971 received on a radio channel and transmitted through the first panel 1910. The second panel 1920 is reconfigurable for transmitting the incident signals 1972 received via the first panel 1910 through the second panel 1920 or reflecting the incident signals 1972 received via the first panel 1910. Some examples may prescribe that the second panel 1920 is reconfigurable for transmitting or reflecting the incident signal based on a characteristic of the incident signal 1971. Such a characteristic may comprise a frequency and/or a polarization of the incident signal 1971. For example, the second panel 1920 may be reconfigurable for transmitting an incident signal having a first frequency and reflecting an incident signal having a second frequency, wherein the second frequency is different from the first frequency. Likewise, the second panel 1920 may be reconfigurable for transmitting an incident signal having a first polarization and reflecting an incident signal having a second polarization, wherein the first polarization is different from, in particular orthogonal to, the second polarization.

In examples, the second panel 1920 may be reconfigurable to be transmissive at a first point in time and to be reflective at a second point in time. In particular, the second panel 1920 may be reconfigurable to be transmissive at a first point in time for an incident signal having a first frequency and/or a first polarization and reflective at a second point in time for an incident signal having the first frequency and/or the first polarization.

Referring to Fig. 19, the incident signal 1971 may be received by the first panel 1910 on a first frequency and a second frequency as indicated with the dash dotted line. The incident signal 1971 is refracted by the first panel 1910. The second panel 1920 receives the refracted incident signal 1972. The portion of the refracted incident signal 1972 having the first frequency is reflected by the second panel 1920. The first panel 1910 may receive the reflected portion 1983 indicated with the dashed line and transmit the refracted reflected portion 1984. On the other hand, the portion of the refracted signal 1972 having the second frequency is transmitted by the second panel 1920. Thus, the portion of the incident signal 1971 having the second frequency passes through the CED 1900 and is transmitted as signal portion 1993 indicated with a dotted line.

In examples, the first panel 1910 may be reconfigurable for changing an angle of refraction of incident signals based on a frequency and/or polarization of the incident signal. For example, with reference to Fig. 19, the angle between signals 1971 and 1984 may depend on the frequency and/or polarization of the signal 1971.

In the example shown in Fig. 19, the reflected signal 1984 has passed through the first panel 1910 twice, wherein the transmitted signal 1993 has passed through the first panel 1910 only once. This may require a larger tuning ratio for the first panel 1910 (i.e., the phase shifting layer).

For quantized beamforming it may be desirable that the CED applies a phase shift of 0 or 180 degrees to the incident signal, i.e. the phase of the emitted (i.e., transmitted or reflected) signal may have to be shifted by 0 or 180 degrees with respect to the incident signal.

Heretofore, the unit cells in the first panel 1910 may provide three phase states with a respective relative through phase of 0, 90 and 180 degrees. In reflective mode, the phase states 0 and 90 degrees may be selected while in the transmissive case the phase states 0 and 180 degrees may be selected.

For example, in reflective mode the incident signal may experience a first phase shift of 90 degree while passing through a unit cell of the first panel 1910, experience a second phase shift of 0 degree while being reflected by the second panel 1920, and experience a third phase shift of 90 degree again while passing through the first panel 1910, which leads to a phase shift of 180 degree between the incident signal and the signal emitted by the CED 1900. In transmissive mode the incident signal may experience a first phase shift of 180 degree while passing through the first panel 1910 and a second phase shift of 0 degree while being transmitted by the second panel 1920, which leads to a phase shift of 180 degree between the incident signal and the signal emitted by the CED 1900.

Fig. 20 shows a further CED 2000. The CED 2000 comprises a first panel 2010, a second panel 2020, and a third panel 2030. The second panel 2020 is arranged between the first panel 2010 and the third panel 2030. The first panel 2010 is reconfigurable for changing an angle of refraction of incident signals 2071 received on a radio channel and transmitted through the first panel 2010. The second panel 2020 is re- configurable for transmitting the incident signals 2072 received via the first panel 2010 through the second panel 2020 or reflecting the incident signals 2072 received via the first panel 2010. The third panel 2030 is reconfigurable for changing an angle of refraction of incident signals 2093 via the second panel 2020. In examples, the incident signal 2071 may be received by the first panel 2010 on a first frequency and a second frequency as indicated with the dash dotted line. The incident signal 2071 is refracted by the first panel 2010. The second panel 2020 receives the refracted incident signal 2072. The portion of the refracted incident signal 2072 having the first frequency is reflected by the second panel 2020 due to the selected stop band defined by the MFSS structure. The first panel 2010 may receive the reflected portion 2083 indicated with the dashed line and transmit as well as refract the refracted reflected portion 2084. The portion of the refracted signal 2072 having the second frequency is transmitted by the second panel 2020. The third panel 2030 receives said portion, refracts and transmit it as signal 2094 indicated with a dotted line. As both the first panel 2010 and the third panel 2030 are reconfigurable for changing an angle of refraction of incident signals, the direction of signals 2084 and 2094 may be configured independently of one another which may be useful in many scenarios.

According to examples, the second panel 2020 may also be configured for attenuating, in particular absorbing, the refracted signal 2072, which is described in more detail with reference to Fig. 21 . Using three panels 2010, 2020, 2030 may relax the requirements concerning the tuning ratio of the first panel 2010. In particular, only two phase states for the unit cells of the first panel 2010 may be required for quantized beamforming. In particular, every incident signal will pass through two times through a refracting (i.e. , phase shifting) panel. In the reflective case, the incident signal passes two times through the first panel 2010. In the transmissive case, the incident signal passes through the first panel 2010 and the third panel 2030.

Another CED 2100 is shown in Fig. 21 . The CED 2100 includes a first panel 2110, a second panel 2120, and a third panel 2130. The second panel 2120 is arranged between the first panel 2110 and the third panel 2130. Like in the examples described before, the first panel 2110 is reconfigurable for changing an angle of refraction of incident signals 2171 received on a radio channel and transmitted through the first panel 2110. Similarly, the second panel 2120 is reconfigurable for transmitting the incident signals 2172 received via the first panel 2110 through the second panel 2120 or reflecting the incident signals 2172 received via the first panel 2110. In the example shown in Fig. 21 , the third panel 2130 is configured to attenuate an incident signal received via the second panel 2120. In an exemplary scenario, the first panel 2110 may receive the incident signal 2171 indicated with a dash dotted line on a first frequency and a second frequency. The first panel 2110 may refract the incident signal 2171 . The second panel 2120 receives the refracted incident signal 2172. The portion of the refracted incident signal 2172 having the first frequency is reflected by the second panel 2120. The first panel 2110 receives the reflected portion 2183 indicated with the dashed line and transmits the refracted reflected portion 2183 as signal 2184. The portion of the refracted signal 2172 having the second frequency is transmitted by the second panel 2120. The third panel 2130 may be configured for attenuating said portion. In some examples, the attenuation may be implemented by switchable loading components applied at each unit cell, e.g., resistors. Thus, the CED 2100 may be considered as a frequency filter only reflecting the portion of the incident signal 2171 having the first frequency.

Fig. 22 illustrates a still further example of a CED 2200. The CED 2200 comprises a first panel 2210 and a second panel 2220, wherein the first panel 2210 is reconfigurable for changing an angle of refraction of incident signals received on a radio channel and transmitted through the first panel 2210, and wherein the second panel 2220 is reconfigurable for transmitting the incident signals received via the first panel 2210 through the second panel 2220 or reflect the incident signals received via the first panel 2210. The first panel 2210 may receive incident signals 2271 , 2281. The incident signals 2271 , 2281 may have different, in particular orthogonal, polarizations. For example, the incident signal 2271 may have a vertical polarization and the incident signal 2281 may have a horizontal polarization. The incident signals 2271 , 2281 are refracted by the first panel 2210. The second panel 2220 receives the refracted incident signals 2272, 2282. The refracted incident signal 2282 having a horizontal polarization may be reflected by the second panel 2220. The first panel 2210 may receive the reflected portion 2283 and transmit the refracted reflected portion 2284. In contrast, the second panel 2220 may transmit the incident signal 2273 having a vertical polarization.

The antenna elements of the panels described hereinbefore could correspond to patches, slots, cross-elements, rings, etc. and built by conductive (e.g., metallic) and dielectric substrates. In particular, the panels may be used for multi-band and/or dually polarized signals according to the structure of the (antenna) elements of the panels. For example, the frequency response of an FSS may be different for different polarizations. If the antenna elements (or unit cells) have a long strip structure, the FSS may be transparent with respect to signals having a polarization that is parallel to the strips but block signals having the orthogonal polarization. On the other hand, a symmetrical structure of the antenna elements (or unit cells) are symmetric in the two orthogonal directions (e.g., squares or circles), the frequency response may be the same for both polarizations.

In the examples depicted in Fig. 17 to 22, the first, second and third panels are depicted as planar panels. However, it is also conceivable that the panels have a different shape. For example, the panels may be curved. In particular, the panels may have a spherical curvature. In some examples, several sub-panels may form a panel. For example, several planar sub-panels may form a spherical panel. Using sub-pan- els may facilitate constructing larger panels. For example, the sub-panels having equal size may be manufactured in large quantity and assembled to larger panels having different sizes.

With reference to Fig. 23, the CEDs described hereinbefore may be particularly useful, if a first communication node (CN) 2301 , for example an access node (AN), is to transmit different signals 2381 , 2391 to further CNs 2302, 2303, for example to two different UEs. The CN 2301 may direct the signal 2371 to the CED 2300 which may reflect and transmit, respectively, the signals 2381 , 2391 to the CNs 2302, 2303. In particular, the CED may enhance coverage for an angle above 180 degrees.

Hence, the proposed CEDs overcome limitations of known reflecting or transmissive CEDs with respect to their coverage area and increase the flexibility in installation of CEDs. In particular, the AoD of the CED may no longer be limited to a half sphere, on either side, depending on if it is a reflective or transmissive RIS. Instead the AoD of the proposed CED may correspond to nearly the full sphere.

In particular, the following examples are disclosed:

EXAMPLE 1.A coverage enhancing device (1900; 2000; 2100; 2200), CED, comprising - a first panel (1910; 2000; 2100; 2200) being reconfigurable for changing an angle of refraction of incident signals (1971 ; 2071 ; 2171 ; 2271 , 2281 ) received on a radio channel and transmitted through the first panel (1910; 2010; 2110; 2210); and

- a second panel (1920; 2020; 2120; 2220) being reconfigurable for transmitting the incident signals received via the first panel (1910; 2010; 2110; 2210) through the second panel (1920; 2020; 2120; 2220) or reflecting the incident signals received via the first panel(1910; 2010; 2110; 2210).

EXAMPLE 2. The CED (1900; 2000; 2100; 2200) of EXAMPLE 1 , wherein the second panel (1920; 2020; 2120; 2220) is reconfigurable for transmitting or reflecting the incident signals (1971 ; 2071 ; 2171 ; 2271 , 2281) based on a characteristic of the incident signal (1971 ; 2071 ; 2171 ; 2271 , 2281 ).

EXAMPLE 3. The CED (1900; 2000; 2100; 2200) of EXAMPLE 1 or 2, wherein the second panel (1920; 2020; 2120; 2220) is reconfigurable for transmitting or reflecting the incident signals (1971 ; 2071 ; 2171 ; 2271 , 2281 ) based on a frequency and/or polarization of the incident signal (1971 ; 2071 ; 2171 ; 2271 , 2281 ).

EXAMPLE 4. The CED (1900; 2000; 2100; 2200) of EXAMPLES 1 to 3, wherein the first panel (1910; 2010; 2110; 2210) is reconfigurable for changing an angle of refraction of incident signals (1971 ; 2071 ; 2171 ; 2271 ) based on a frequency and/or polarization of the incident signal (1971 ; 2071 ; 2171 ; 2271 , 2281 ).

EXAMPLE 5. The CED (2000; 2100) of any one of EXAMPLES 1 to 4, further comprising a third panel (2030; 2130), wherein the second panel (2020; 2120) is provided between the first panel (2010; 2110) and the third panel (2030; 2130).

EXAMPLE 6. The CED (2000) of EXAMPLE 5, wherein the third panel (2030) is reconfigurable for changing an angle of refraction of incident signals (2083) received via the second panel (2020).

EXAMPLE 7. The CED (2100) of EXAMPLE 5 or 6, wherein the third panel (2130) is reconfigurable for attenuating an incident signal received via the second panel (2120).

EXAMPLE 8. The CED (1900; 2000; 2100; 2200) of any one of EXAMPLES 1 to 7, wherein the first panel (1910; 2010; 2110; 2210) and/or the second panel (1920; 2020; 2120; 2220) and/or the third panel (2030; 2130; 2230) are electronically reconfigurable.

EXAMPLE 9. The CED (1900; 2000; 2100; 2200) of any one of EXAMPLES 1 to 8, wherein the first panel (1910; 2010; 2110; 2210) and/or the second panel (1920; 2020; 2120; 2220) and/or the third panel (2030; 2130) are mechanically reconfigurable.

EXAMPLE 10. The CED (1900; 2000; 2100; 2200) of any one of EXAMPLES 1 to 9, wherein the first panel (1910; 2010; 2110; 2210) and/or the third panel (2030; 2130) comprise a phase shifting layer.

EXAMPLE 11 . The CED (1900; 2000; 2100; 2200) of any one of EXAMPLES 1 to 10, wherein at least one of the first panel (1910; 2010; 2110; 2210) and/or the second panel (1920; 2020; 2120; 2220) and/or the third panel (2030; 2130) comprises two or more sub-panels.

EXAMPLE 12. The CED (1900; 2000; 2100; 2200) of EXAMPLE 11 , wherein at least some of the sub-panels are curved.

EXAMPLE 13. The CED (1900; 2000; 2100; 2200) of EXAMPLE 11 or 12, wherein at least some of the sub-panels are planar.

Although the invention has been shown and described with respect to certain preferred examples, equivalents and modifications will occur to others skilled in the art upon the reading and understanding of the specification. The present invention includes all such equivalents and modifications and is limited only by the scope of the appended claims.