Introduction
Future wireless systems must support uninterrupted operation of massive Internet of Things networks, providing reliable energy supply for low-power wireless devices. Radio frequency (RF) wireless power transfer (WPT) is a promising solution to achieve this goal. The end-to-end power transfer efficiency of an RF-WPT system is inherently low, calling for efficient techniques such as waveform optimization, energy beamforming (EB), and distributed antenna systems [1], [2].
EB can focus energy beams toward one or more energy harvesting (EH) devices at the same time. The EB flexibility and potential gains are determined by the transmitter architecture. Although a fully-digital (FD) structure provides the highest flexibility, it requires a large number of RF chains, including power amplifiers (PAs), resulting in high complexity, cost, and power consumption. The analog architecture alternatives incur much lower power consumption and cost by reducing the number of RF chains, and utilizing only analog circuits as phase shifters. However, they may not offer sufficient flexibility, thus, hybrid architectures have been introduced to combine digital and analog approaches and provide beamforming capability with a trade-off between complexity/cost and flexibility [3].
In hybrid architectures, RF chains and phase shifters can be connected using, e.g., the fully-connected network, where all RF chains are connected to all phase shifters. However, since the complexity and losses of analog circuits affect scalability and performance, simpler array of subarrays (AoSA)-based architectures are used [4]. Another technology that avoids analog circuits is the dynamic metasurface antenna (DMA), a group of configurable metamaterial elements placed on a set of waveguides [5]. Unlike reflective surfaces [6], where there is no correlation between the phase shift and gain introduced by each reflecting element, the modeling of DMA metamaterial elements is more involved, making their corresponding design problems different in general. DMA is utilized in [7] in near-field WPT to maximize the sum-harvested energy considering linear RF-to-direct current (DC) power conversion. However, in practice, there are extensive non-linearities in the EH device. Thus, a linear power conversion model cannot represent the amount of harvested power in a practical EH device. Herein, we also consider a DMA transmitter architecture to charge multiple single-antenna EH devices but utilize a practical alternative, which assumes that each device has a specific DC EH requirement and that the corresponding RF power requirement can be obtained by the nonlinear EH relationship of the device. Then, providing that specific RF power required by the device leads to meeting the EH requirements.
Our main contributions are: i) we optimize the EB aiming to minimize the transmit power while meeting the EH requirements of all receivers; ii) we provide a frequency-dependant model for the propagation characteristics of the DMA, which was previously overlooked; iii) we propose an efficient EB solution relying on semi-definite programming (SDP) and alternating optimization, which converges in polynomial time; iv) we show that DMA outperforms the FD structure in terms of minimum transmit power for meeting the RF power requirements, and that the transmit power reduces with increasing the antenna array size, while it increases with the operating frequency in the DMA architecture. Next, Section II presents the system model and the optimization problem. In Section III, we discuss the proposed solution, while Section IV provides numerical results. Finally, Section V concludes this letter.
Notations: Bold lowercase (upper-case) letters represent vectors (matrices),
System Model & Problem Formulation
We consider a multi-antenna RF-WPT system where a transmitter with a DMA uniform square array charges
Channel Model: We consider the near-field wireless channel. The
-th user, located at a distance{k} from the transmitter, lies in the near-field propagation region ifr_{k} , whered_{fs} < r_{k} < d_{fr} andd_{fs} = \sqrt [{3}]{\frac {D^{4}}{8\lambda }} are the Fresnel and Fraunhofer distance, respectively. Moreover,d_{fr} = {}\frac {2D^{2}}{\lambda } is the wavelength,\lambda is the antenna diameter, andD = \sqrt {2}L is the antenna length. The location of the{L} th element in the{l} th waveguide is{i} ,\mathbf {p}_{i,l} = [x_{i,l}, y_{i,l}, z_{i,l}]^{T} andi = 1,2,\ldots,N_{v} , whilel = 1,2,\ldots, N_{h} denotes the\mathbf {p}_{k} th device location.{k} The channel coefficient between user
and the{k} th metamaterial element in the{l} th waveguide is given by [8]{i} where\begin{equation*} \gamma _{i,l,k} = A_{i,l,k} e^{-j2\pi d_{i,l,k}/\lambda }, \tag{1}\end{equation*} View Source\begin{equation*} \gamma _{i,l,k} = A_{i,l,k} e^{-j2\pi d_{i,l,k}/\lambda }, \tag{1}\end{equation*}
constitutes the phase shift introduced by the propagation distance, and2\pi d_{i,l,k}/\lambda is the Euclidean distance between the element and the user. Additionally, the channel gain coefficient isd_{i,l,k} = |\mathbf {p}_{k} - \mathbf {p}_{i,l}| where\begin{equation*} A_{i,l,k} = \sqrt {F\left ({\theta _{i,l,k},\phi _{i,l,k}}\right)}\frac {\lambda }{4\pi d_{i,l,k}}, \tag{2}\end{equation*} View Source\begin{equation*} A_{i,l,k} = \sqrt {F\left ({\theta _{i,l,k},\phi _{i,l,k}}\right)}\frac {\lambda }{4\pi d_{i,l,k}}, \tag{2}\end{equation*}
and\theta _{i,l,k} are the elevation and azimuth angle, respectively, while\phi _{i,l,k} denotes the radiation profile of each element. We assume that the latter is given by [9]F(\theta _{i,l,k},\phi _{i,l,k}) where\begin{align*} F\left ({\theta _{i,l,k},\phi _{i,l,k}}\right) = \begin{cases} G_{t}{\cos {\left ({\theta _{i,l,k}}\right)}}^{\frac {G_{t}}{2}-1}, & \theta _{i,l,k} \in \left [{0,\frac {\pi }{2}}\right], \\ 0, & \mathrm {otherwise}, \end{cases} \tag{3}\end{align*} View Source\begin{align*} F\left ({\theta _{i,l,k},\phi _{i,l,k}}\right) = \begin{cases} G_{t}{\cos {\left ({\theta _{i,l,k}}\right)}}^{\frac {G_{t}}{2}-1}, & \theta _{i,l,k} \in \left [{0,\frac {\pi }{2}}\right], \\ 0, & \mathrm {otherwise}, \end{cases} \tag{3}\end{align*}
andG_{t} = 2(b+1) are the transmit antenna gain and the boresight gain, respectively. Now, let us define{b} as the\boldsymbol {\gamma }_{k} = \begin{bmatrix} \gamma _{1,1,k},\gamma _{1,2,k}, \ldots, \gamma _{N_{v},N_{h},k}\end{bmatrix}^{T} -dimensional channel coefficients vector between the user{N} and the elements of the DMA. Note that under the conventional far-field conditions, the channel coefficient can be represented as{k} , whereA_{k} e^{-j\psi _{i,l,k}} is only determined by the distance between the user and the antenna array, andA_{k} by the spatial direction of the user and the spacing between the antenna elements.\psi _{i,l,k} DMA Model: In practice, microstrip lines are usually utilized as the DMA waveguides. The propagation coefficient for the
th element in the{l} th microstrip is given by [10]{i} where\begin{equation*} h_{i, l} = e^{-(l-1)d_{l}\left ({\alpha _{i}+j\beta _{i}}\right)}, \tag{4}\end{equation*} View Source\begin{equation*} h_{i, l} = e^{-(l-1)d_{l}\left ({\alpha _{i}+j\beta _{i}}\right)}, \tag{4}\end{equation*}
,d_{l} , and\alpha _{i} are the inter-element distance, waveguide attenuation coefficient, and propagation constant, respectively. We assume that all microstrips are of the same type;\beta _{i} and\beta _{i} = \beta . In general, the frequency of the system affects the signal propagation inside a waveguide, and hence, the propagation model must capture the frequency-dependent effects.\alpha _{i} = \alpha The microstrip line comprises a conductor of width
, printed on a dielectric substrate with thickness\upsilon and dielectric constant\zeta . Thus, the effective dielectric constant at DC is [11]\epsilon _{r} The effective dielectric constant at frequency\begin{equation*} \epsilon ^{\prime }_{e} = {\left ({\epsilon _{r}+1}\right)}/{2} + {\left ({\epsilon _{r}-1}\right)}/{2}{\sqrt {1 + 12{\zeta }/{\upsilon }}}. \tag{5}\end{equation*} View Source\begin{equation*} \epsilon ^{\prime }_{e} = {\left ({\epsilon _{r}+1}\right)}/{2} + {\left ({\epsilon _{r}-1}\right)}/{2}{\sqrt {1 + 12{\zeta }/{\upsilon }}}. \tag{5}\end{equation*}
is [11]{f} where\begin{equation*} \epsilon _{e}^{f} = \epsilon _{r} - {\bigl (\epsilon _{r}-\epsilon ^{\prime }_{e}\bigr)}/{\bigl (1 + G(f)\bigr)}, \tag{6}\end{equation*} View Source\begin{equation*} \epsilon _{e}^{f} = \epsilon _{r} - {\bigl (\epsilon _{r}-\epsilon ^{\prime }_{e}\bigr)}/{\bigl (1 + G(f)\bigr)}, \tag{6}\end{equation*}
, andG(f) = {(0.6 + 0.009Z_{0})f^{2}}/{ (Z_{0}/8\pi \zeta)^{2}} is the characteristic impedance of the microstrip, while both cases in (7) are approximately equal for\begin{align*} Z_{0} = \begin{cases} 60 \ln {\left ({8\zeta /\upsilon + \upsilon /(4\zeta)}\right)}/\sqrt {\epsilon ^{f}_{e}}, & \upsilon \leq \zeta, \\ \frac {120\pi /\sqrt {\epsilon ^{f}_{e}}}{\upsilon /\zeta + 1.393+0.667\ln {\left ({\upsilon /\zeta + 1.444}\right)}}, & \upsilon \geq \zeta,\end{cases} \tag{7}\end{align*} View Source\begin{align*} Z_{0} = \begin{cases} 60 \ln {\left ({8\zeta /\upsilon + \upsilon /(4\zeta)}\right)}/\sqrt {\epsilon ^{f}_{e}}, & \upsilon \leq \zeta, \\ \frac {120\pi /\sqrt {\epsilon ^{f}_{e}}}{\upsilon /\zeta + 1.393+0.667\ln {\left ({\upsilon /\zeta + 1.444}\right)}}, & \upsilon \geq \zeta,\end{cases} \tag{7}\end{align*}
. Thereby,{\upsilon }/{\zeta } = 1 is the microstrip propagation constant and the attenuation due to the dielectric loss is given by\beta = ({2\pi }/{\lambda })\sqrt {\epsilon ^{f}_{e}} where\begin{equation*} \alpha _{d} = \pi \epsilon _{r}\left ({\epsilon ^{f}_{e} - 1}\right)\varrho /\Bigl ({\lambda \sqrt {\epsilon _{e}^{f}}\left ({\epsilon _{r} -1}\right)}\Bigr), \tag{8}\end{equation*} View Source\begin{equation*} \alpha _{d} = \pi \epsilon _{r}\left ({\epsilon ^{f}_{e} - 1}\right)\varrho /\Bigl ({\lambda \sqrt {\epsilon _{e}^{f}}\left ({\epsilon _{r} -1}\right)}\Bigr), \tag{8}\end{equation*}
is the loss tangent of the dielectric. Moreover,\varrho is the approximate attenuation due to the conductor loss and\alpha _{c} = R_{s}/Z_{0}\upsilon is the conductor surface resistivity withR_{s} = \sqrt {2\pi f \mu _{0}/2\sigma } and\sigma being the conductivity and the free space permeability, respectively. Finally, the microstrip line attenuation coefficient is modeled as\mu _{0} . Table I lists some microstrip materials with their main characteristics.\alpha = \alpha _{d} + \alpha _{c} Next,
is the microstrip propagation diagonal matrix with\mathbf {H} \in \mathbb {C}^{N \times N} being theh_{i,l} th column and((i-1)N_{h} + l) th row element. The Lorentzian-constrained phase model is utilized to capture the dependency between the elements amplitude and phase [10], so that the frequency response of the((i-1)N_{h} + l) th element in the{l} th microstrip is{i} Additionally,\begin{equation*} q_{i, l} \in \mathcal {Q} = \Big \{{ \left ({j+e^{j \phi _{i,l}}}\right)}/{2} \Big | \phi _{i,l} \in \left [{0,2 \pi }\right]\Big \}, \quad \forall i, l. \tag{9}\end{equation*} View Source\begin{equation*} q_{i, l} \in \mathcal {Q} = \Big \{{ \left ({j+e^{j \phi _{i,l}}}\right)}/{2} \Big | \phi _{i,l} \in \left [{0,2 \pi }\right]\Big \}, \quad \forall i, l. \tag{9}\end{equation*}
is the matrix containing the configurable weights of the metamaterial elements [8], i.e.,\mathbf {Q} \in \mathbb {C}^{N \times N_{d}} \begin{align*} \mathbf {Q}_{(i-1) N_{h}+l, n}= \begin{cases}q_{i, l}, & i=n, \\ 0, & i \neq n.\end{cases} \tag{10}\end{align*} View Source\begin{align*} \mathbf {Q}_{(i-1) N_{h}+l, n}= \begin{cases}q_{i, l}, & i=n, \\ 0, & i \neq n.\end{cases} \tag{10}\end{align*}
Signal Model: Let
be thex_{m} th energy symbol at the input of the digital precoder where{m} andm = 1,\ldots, M . Also, the energy symbols are independent and normalized such thatM = \min (N_{v}, K) and\mathbb {E}\{x_{n}^{H} x_{r}\} = 0 , while\mathbb {E}\{x_{n}^{H} x_{n}\} = 1 is the\mathbf {w}_{m} -dimensional precoding vector forN_{v} . Thus, the transmit signal isx_{m} , while the transmit power is\mathbf {s} = \sum _{m = 1}^{M} \mathbf {H Q w}_{m}x_{m} . The received energy signal at theP_{Tx} = \mathbb {E}_{x}\{\mathbf {s}^{H} \mathbf {s}\} = \sum _{m = 1}^{M} |\mathbf {H Q w}_{m}|^{2} th EH device is given by{k} , while the corresponding received RF power isy_{k} = \boldsymbol {\gamma }_{k}^{H}\mathbf {s} \begin{equation*} P_{Rx}^{k} = \mathbb {E}_{x}\bigl \{y_{k}^{H} y_{k}\bigr \} = \sum _{m = 1}^{M} |\boldsymbol {\gamma }_{k}^{H} \mathbf {H Q w}_{m}|^{2}. \tag{11}\end{equation*} View Source\begin{equation*} P_{Rx}^{k} = \mathbb {E}_{x}\bigl \{y_{k}^{H} y_{k}\bigr \} = \sum _{m = 1}^{M} |\boldsymbol {\gamma }_{k}^{H} \mathbf {H Q w}_{m}|^{2}. \tag{11}\end{equation*}
A. Problem Formulation
In practical WPT systems, the transmit power influences largely the total power consumption. Indeed, the transmit power determines the power consumption of the PAs, which are the most power-hungry system components. Motivated by this, our objective is to minimize the transmit power while satisfying the users’ RF power requirements. Assuming the location of the users is known, the optimization problem is \begin{align*} {{\mathbf { {P1}}}:} \quad \mathop {\mathrm {minimize}} _{\mathbf {Q}, \{\mathbf {w}_{m}\},\forall m}&\sum _{m = 1}^{M} |\mathbf {H Q w}_{m}|^{2} \tag{12a}\\ {\mathrm {subject to}}&\sum _{m = 1}^{M} |\boldsymbol {\gamma }_{k}^{H} \mathbf {H Q w}_{m}|^{2} \geq \delta _{k}, \quad \forall k, \tag{12b}\\&(10),\quad q_{i,l} \in \mathcal {Q}, \quad \forall i,l. \tag{12c}\end{align*}
Proposed EB Optimization Solution
A. Optimal Digital Precoders With Fixed Q
Let us fix Q and rewrite (11) as \begin{align*} P_{Rx}^{k}=&\sum _{m = 1}^{M}\left ({\left ({\boldsymbol {\gamma }_{k}^{H} \mathbf {H Q}}\right) \mathbf {w}_{m}}\right)^{H} \left ({\boldsymbol {\gamma }_{k}^{H} \mathbf {H Q}}\right) \mathbf {w}_{m} \\=&\mathrm {Tr}\left({\left[{\sum _{m = 1}^{M} \mathbf {w}_{m} \mathbf {w}_{m}^{H}}\right] \mathbf {b}_{k} \mathbf {b}_{k}^{H} }\right) = \mathrm {Tr}\bigl (\mathbf {W}\mathbf {B}_{k}\bigr), \tag{13}\end{align*}
\begin{align*} {{\mathbf { {P2}}}:} \quad \mathop {\mathrm {minimize}} _{\mathbf {W}}&\mathrm {Tr}\bigl (\mathbf {W}\mathbf {F}\bigr) \tag{14a}\\ {\mathrm {subject to}}&\mathrm {Tr}\bigl (\mathbf {W}\mathbf {B}_{k}\bigr) \geq \delta _{k}, \quad \forall k, \tag{14b}\\&\mathbf {W} \succeq \mathbf {0}, \tag{14c}\end{align*}
B. Suboptimal Q With Fixed Digital Precoders
Now, let us fix \begin{equation*} |\boldsymbol {\gamma }_{k}^{H} \mathbf {H Q w}_{m}|^{2}=\left |{\left ({\mathbf {w}_{m}^{T} \otimes (\boldsymbol {\gamma }_{k}^{H} \mathbf {H})}\right) \mathbf {q}}\right |^{2} = |\mathbf {z}_{m,k}^{H} \mathbf {q}|^{2}, \tag{15}\end{equation*}
\begin{align*} P_{Rx}^{k}{\overset{(a)}{=}}&\sum _{m = 1}^{M} \left ({\mathbf {z}_{m,k}^{H} \mathbf {q}}\right)^{H} \left ({\mathbf {z}_{m,k} \mathbf {q}}\right) \\{\overset{(b)}{=}}&\hat {\mathbf {q}}^{H} \left[{\sum _{m = 1}^{M} \hat {\mathbf {z}}_{m,k} \hat {\mathbf {z}}_{m,k}^{H} }\right] \hat {\mathbf {q}} {\overset{(c)}{=}} \mathrm {Tr}\bigl (\widetilde {\mathbf {Z}}_{m,k} \widetilde {\mathbf {Q}}\bigr), \tag{16}\end{align*}
\begin{align*} {{\mathbf { {P3}}}:} \quad \mathop {\mathrm {maximize}} _{\widetilde {\mathbf {Q}}}&\min _{k} \mathrm {Tr}\bigl (\widetilde {\mathbf {Z}}_{m,k} \widetilde {\mathbf {Q}}\bigr) \tag{17a}\\ {\mathrm {subject to}}&\widetilde {\mathbf {Q}} \succeq \mathbf {0}, \tag{17b}\\&q_{i,l} \in \mathcal {Q}, \quad \forall i,l, \tag{17c}\\&\mathrm {rank}\left ({\widetilde {\mathbf {Q}}}\right) = 1. \tag{17d}\end{align*}
Problem (17) is still difficult to solve due to the Lorentzian and rank constraints. Therefore, we relax the problem as \begin{align*} {{\mathbf { {P4}}}:} \quad \mathop {\mathrm {maximize}} _{\widetilde {\mathbf {Q}}, t}&t \tag{18a}\\ {\mathrm {subject to}} \quad&t \leq \mathrm {Tr}\bigl (\widetilde {\mathbf {Z}}_{m,k} \widetilde {\mathbf {Q}}\bigr), \quad \forall k, \tag{18b}\\&\mathrm {Tr}\bigl (\widetilde {\mathbf {Q}}\bigr) \leq N, \tag{18c}\\&\widetilde {\mathbf {Q}} \succeq \mathbf {0}, \tag{18d}\end{align*}
\begin{equation*} \phi ^{\star }_{i,l} = \min _{\phi _{i,l}} \quad \Big |\left ({{j + e^{j{\phi }_{i,l}}}}\right)/{2} - q^{\prime }_{i,l}\Big |, \quad \forall i,l, \tag{19}\end{equation*}
C. Overall Algorithm
Algorithm 1 summarizes the overall EB optimization solution for DMA-assisted RF-WPT system. At first, a random Lorentzian constrained frequency response is generated for each metamaterial element, and Q is constructed. Then, a local optimum solution is obtained for P3 through lines 6-7, followed by finding the optimal precoders by solving P2. Moreover, the solution is updated in lines 10–12 if the transmit power decreases. The above process is repeated iteratively until there is no improvement in the solution for a maximum stall counter or a maximum number of iterations is reached.
Algorithm 1 EB Optimization for DMA-Based RF-WPT
Input:
Initialize:
Solve P2 to obtain
repeat
Solve P4 to obtain
Obtain
Solve P2 to obtain
if
end if
until
The number of variables in (14) and (18) is
Numerical Results
We consider a 100 m2 indoor area with a transmitter at the center of the ceiling with a 3 m height. The users are randomly distributed over 30 realizations and
We refer to the proposed method as EB-ASD, while particle swarm optimization (PSO) [14] with 1000 iterations and 100 particles per iteration is used as a benchmark. The DMA performance is compared to the FD structure with an inter-element distance of
Fig. 3(c) illustrates the simulation results over different operating frequencies. It is observed that the proposed approach outperforms PSO. The reason is that the problem space is large, especially in high-frequency regime, and thus, PSO needs much more time and particles to perform well. For a multi-user setup, maximum ratio transmission (MRT)-based [15] precoders can be utilized to derive a lower bound for the transmit power of the FD structure. Hereby, one may expect that the received RF power scales with
(a) Average transmit power (top) and (b) average gain and the number of antenna elements (bottom), as a function of frequency for fixed
The influence of the antenna length is illustrated in Fig. 4(c). The discussions of the previous case are also applicable here. As expected, the transmit power decreases with antenna length in both architectures since the average effective gain increases. Additionally, DMA outperforms FD in this specific frequency but the situation may change at significantly higher frequencies, as DMA losses become considerably large. The proposed approach also outperforms PSO in this configuration.
(a) Average transmit power (top) and (b) average gain and the number of antenna elements (bottom), as a function of antenna length for fixed
Conclusion
We considered a multi-user RF-WPT system with a DMA-based architecture. Moreover, we proposed an EB design aiming to minimize the transmit power while meeting the users’ EH requirements. The solution, using SDP and alternating optimization, outperformed PSO, and our results also reveal that a DMA transmitter is preferred to the FD alternative. It was observed that although increasing the system frequency does not affect the FD architecture performance, it increases the required transmit power for the DMA-assisted system since the effective gain decreases. Meanwhile, increasing the number of elements by utilizing a larger antenna array may lead to significant performance gains in both architectures.