INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 340
Design and Optimisation of a 2.4 GHz Radio-Frequency Energy
Harvesting System for Low-Power Internet of Things (IoT)
Applications
John Nyamekye Ansah., John Kojo Annan., Shiphrah Ohene Adu
Department of Electrical and Electronic Engineering, University of Mines and Technology, Tarkwa-
Ghana
ABSTRACT
This paper presents the design and performance evaluation of a highly efficient 2.4 GHz Radio-Frequency
Energy Harvesting (RF-EH) system specifically modified to meet the energy requirements of low-power Internet
of Things (IoT) applications. This system incorporates a high-gain 4x4 Microstrip Patch Antenna (MPA) Array,
a single-stub Matching Network (MN) and a 5-stage RF-to-DC rectifier using the Cockcroft-Walton Voltage
Multiplier (CW-VM) topology. The antenna is mounted on Rogers RT5880 substrate with a thickness of 1.588
mm and a relative permittivity of 2.2, and fed by a 50-Ω microstrip feed line. The antenna was designed using
Computer Simulation Technology (CST) Studio Suite 2019, while the rectifier and MN were implemented in
Keysight Advanced Design System (ADS) 2022. The antenna achieved a high gain of 19.29 dBi and a directivity
of 20.01 dBi, with radiation and total efficiencies exceeding 84 %. The antenna demonstrated a highly directive
E-Plane radiation pattern with a narrow beamwidth of 19.4
° and a side lobe level of -13.8 dB, indicating a more
focused reception of RF signals and a good suppression of unwanted radiation. The rectifier achieved a peak
Power Conversion Efficiency (PCE) of 89.913 % and a peak output voltage of 16.424 V at an input power of
17.5 dBm. More importantly, low-input powers ranging from -20 dBm to 0 dBm produced usable DC output
voltages from 0.456 V to 4.144 V, respectively, demonstrating strong suitability for IoT applications operating
under limited RF conditions. These results demonstrated the system’s potential for integration into large-scale
indoor and outdoor IoT networks. The proposed system supports sustainable, maintenance-free and battery-less
deployments, advancing the development of autonomous wireless systems.
Keywords: Radio-Frequency, Energy Harvesting, Antenna, Matching Network, Rectifier Circuit, IoT
Applications
INTRODUCTION
The rapid proliferation of wireless sensors and Internet of Things (IoT) devices in recent times has intensified
energy demand challenges. Most IoT nodes predominantly rely on battery power, necessitating frequent
replacement and maintenance: an impractical and costly approach for large-scale or remote deployments. Their
limited lifespan restricts scalability, while safety concerns, such as the risk of overheating and leakage, pose
significant threats in industrial or high-temperature environments like agricultural farms and greenhouses [1].
Moreover, batteries contain toxic chemicals such as lead, lithium, and cobalt, which pose serious risks to the
soil, aquatic life and human health if not properly disposed of. With IoT deployments expected to exceed 30
billion devices by 2030, the ecological consequences of large-scale battery use are projected to escalate. This
points to the pressing need for a sustainable, eco-friendly and battery-less energy source to power IoT systems
efficiently and at a low cost [2].
The realisation of autonomous and battery-less IoT systems can be effectively achieved through Energy
Harvesting (EH) [3]. EH involves the collection of available ambient energy sources such as solar, thermal,
wind, sound, vibration and Radio-Frequency (RF) signals. Among these, Radio-Frequency Energy Harvesting
(RF-EH) offers a promising solution, primarily due to the global deployment of 5
th
Generation (5G) and ubiquity
wireless infrastructure, such as mobile cell towers, Wi-Fi routers, FM radio and television (TV) transmitters.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 341
This persistent availability of RF energy makes it possible to harvest and convert ambient signals into usable
power to drive low-power devices [4]. Hence, RF-EH presents a more favourable and sustainable solution for
powering the next generation of wireless sensors and IoT devices [5].
The process of scavenging of electromagnetic (EM) energy from ambient RF bands is called RF-EH. RF-EH
can be classified into near-field (NF) and far-field (FF) techniques, depending on propagation distance. Near-
Field Energy Harvesting (NF-EH) transfers energy over short distances using inductive coupling, where coils
generate a magnetic field to induce current. This technique is commonly used in wireless charging applications
such as electric vehicles, smartphones and medical implants, but its efficiency decreases with increasing coil
separation [6]. In contrast, Far-Field Energy Harvesting (FF-EH) employs antennas to intercept EM radiations
which travel over long distances and is a widely adopted method for ambient RF-EH due to the ever-present
availability of RF signals. However, FF-EH suffers from reduced input power density as a result of Free-Space
Path Loss (FSPL), which increases with distance. Since it is impossible to capture all the transmitted RF power
at the receiver end, improving RF-to-DC power conversion efficiency (PCE) under low input power conditions
remains a key research challenge in RF-EH designs [7].
Model of Radio-Frequency Energy Harvesting System
A block diagram of the RF-EH system is presented in Figure 1. It consists of (a) a receiver antenna for collecting
RF power from the target frequency band using the FF method; (b) a Matching Network (MN) to convey the
received signal to a rectifier circuit by minimising reflection losses; (c) a rectifier circuit to convert the input
alternating current (AC) signal into direct current (DC) signal and also acts as a voltage boosting block hence
improving the output voltage at the end; (d) a DC pass filter to further remove higher order harmonics of AC
signal which may escape rectification and (e) a DC-to-DC converter to maintain a stable DC output voltage
irrespective of fluctuations in the input power or the load to serve low-power devices efficiently [8].
Figure 1. Block diagram of Radio-Frequency Energy Harvesting System
Recent studies have shown that RF signals are densely concentrated within the 1700-2650 MHz range, covering
numerous communication bands, such as GSM-1800 (1805-1880 MHz), UMTS (1930-1990 MHz) (2110-2170
MHz), Wi-Fi/Bluetooth (2400-2483.5 MHz) and LTE-2600 (2620-2690 MHz). Hence, a substantial amount of
RF energy is available within these bands, which can be effectively captured and converted into a usable DC
signal [8].
However, since RF power density decreases with increasing propagation distance, the received signal is usually
weak at the harvesting node. Therefore, an RF-EH system needs to employ a high-efficient antenna for effective
energy capture, a well-optimised MN to maximise power transfer (PT) to the rectifier and a high-performance
rectifier with enhanced PCE and DC output voltage. These components collectively ensure the sustainable
operation of low-power IoT devices while minimising the reliance on standard batteries [7].
Receiver Antenna
The antenna acts as the initial interface between ambient RF signals and the EH circuitry. Single-band antennas
achieve high efficiency at a specific frequency but fail when the target frequency is unavailable, leading to weak
received input power. Broadband antennas can harvest energy across a wide frequency range but suffer from
low efficiency. Multi-band antennas, on the other hand, provide a balanced solution by targeting selected
frequencies, thus improving both efficiency and output power. Antennas may also be directional or
omnidirectional: directional antennas offer high gain and efficient energy capture from fixed RF sources, but
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 342
require precise alignment, whereas, omnidirectional antennas, though lower in gain, are mostly preferred when
the direction of incoming signals is unknown [9].
The Microstrip Patch Antenna (MPA) is widely used in wireless communication and RF-EH, due to its
compactness, lightweight, low cost, robustness and ease of fabrication using printed circuit technology.
However, despite the suitability of MPAs for RF-EH, they suffer from several limitations when implemented as
a single-patch antenna. These include low efficiency, narrow bandwidth, spurious radiation, poor power-
handling capacity and low gain [7].
Matching Network
In the RF-EH system, frequent high impedance variations often arise from the non-linear behaviour of rectifier
components such as diodes or transistors [5]. To minimise reflection losses and enhance PT between the antenna
and the rectifier, a Matching Network is introduced. MNs can be realised using either lumped elements or
distributed elements. Lumped-Element Matching Networks (LE-MNs) use passive components like inductors,
capacitors, resistors and transformers. Commonly used topologies include L-matching, Pi (π)-matching and T-
matching networks, which are effective at low frequencies (< 500 MHz), where component dimensions remain
small relative to the signal wavelength.
At high frequencies (> 500 MHz), Distributed-Element Matching Networks (DE-MNs) are preferred since
lumped elements become impractical due to parasitic effects and difficulties in achieving precise capacitance or
inductance values. They employ transmission line (TL) sections such as microstrip lines or coplanar waveguides
to perform impedance matching (IM). DE-MNs are primarily classified into three types: quarter-wave
transformers, stub (single or double) MNs and multi-section TL transformers. In DE-MNs, impedance
transformation (IT) is achieved by adjusting the length and width of the TL sections [10].
Rectifier Circuit
The rectifier circuit is a key component of RF-EH systems, converting weak AC signals from the MN into DC
signals and boosting it to provide a sufficient output. Rectifiers are generally classified as half-wave or full-
wave. Half-wave rectifiers, consisting of a single diode connected in series or parallel with the output load, are
simple and low-cost but suffer from high voltage ripple and poor PCE, making them unsuitable for sensitive
applications. Full-wave rectifiers are classified as Dickson Charge Pump (DCP), Cockcroft-Walton (CW) and
Cross-Coupled Differential-Drive (CCDD) rectifiers. A single-stage CW rectifier is also called a Greinacher
Voltage Doubler (GVD). Among the common voltage multiplier (VM) topologies, the CW multiplier is well-
suited for low-power applications that require a high DC output. Its simple, passive, scalable and clock-free
architecture allows for easy addition of multiple stages to increase the output voltage under low input power
conditions. Unlike DCP, which requires clock signals, or CCDD, which suffers from scalability and layout
sensitivity, the CW rectifier remains the preferred choice for RF-EH systems [5, 10].
In conclusion, RF-EH offers a promising approach toward autonomous, battery-less IoT systems by leveraging
ambient RF energy and optimising key components such as antennas, MNs and rectifiers. This approach may
significantly reduce dependence on conventional batteries, lowering cost and supporting sustainability.
LITERATURE REVIEW
Significant research work has been done on RF-EH systems operating at the Industrial Scientific and Medical
(ISM) band. The authors in [8] presented a small dual-band antenna with gains of 1.6 dBi and 3.95 dBi at 2.4
GHz and 5.2 GHz, respectively. The two-ring resonators reported in [3] measured a peak gain of about 5 dBi
between 1.68-2.80 GHz, with a 50 % fractional bandwidth. A dual-band MPA was reported in [9]. It achieved
low gains of 0.5102 dBi and 1.175 dBi, with poor radiation efficiencies at 2.4 GHz and 5.8 GHz, respectively.
Three models of MPA comprising 1x1, 2x2 and 4x4 were reported in [12]. The 4x4 model recorded the highest
gain of 13.05 dBi, operating effectively within 2.32-2.40 GHz. However, its large physical size poses integration
challenges in miniaturised systems. The antenna proposed in [13] was optimised using machine learning
algorithms to reduce simulation time and accelerate the design process. The antenna with omnidirectional
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 343
radiation properties operated effectively between 1.72-3.02 GHz and 3.47-4.79 GHz, covering mobile frequency
and Wireless Local Area Network (WLAN) bands. The authors in [14] presented a 2x2 MPA array to harness
mobile frequency signals from the GSM-1800 band. It achieved a gain of 9.2 dBi with a good return loss value
of -25 dB. Four models of triangular MPA arrays comprising 1x2, 2x2, 2x3 and 2x4 were designed in [15]. The
2x2 model outperformed both the 2x3 and 2x4 arrays with a gain of 11 dBi, even though it had fewer patch
elements. A three-element patch array was presented in [7] for RF-EH. It measured a gain of 11.24 dBi with
radiation and total efficiencies exceeding 89 %. The authors in [16] proposed a seven-element MPA array with
a gain of 15.8 dBi. The 2.4 GHz MPA array reported in [17] achieved a gain of 17.5 dBi.
A 2.45 GHz compact-sized rectenna (antenna + rectifier) was reported in [18]. The system employed an L-type
MN and single-stage GVD, all fabricated on an FR-4 substrate. The rectifier comprised a pair of HSMS-2850
Schottky diodes chosen for their low turn-on voltage and fast switching speed at high frequencies. The antenna
achieved a gain of 1.3 dBi, while the output voltage and PCE of the rectifier recorded 2.9 V and 75 %,
respectively, at 10 dBm. The single-stage RF-to-DC converter presented in [6] achieved an output voltage of 21
mV at -18 dBm. The rectenna proposed in [19] comprised a 2x1 array antenna with an L-type MN selected for
compactness. A rectenna operating at 2.45 GHz was proposed in [20] for RF-EH applications. The design
recorded a gain of 7.31 dBi, with maximum output voltage and PCE of 5.2 V and 64 % respectively, at 0 dBm.
The rectenna proposed in [21] comprised a 1x2 MPA and a seven-stage CW rectifier with an L-type MN. The
antenna achieved a gain of 7.03 dBi with an output voltage of 12 V and PCE of 75 % at 20 dBm.
Previously reported RF-EH designs face significant challenges, notably low antenna gain [3, 8, 9, 13, 18] and
poor radiation efficiency [12, 14, 15, 17], both of which hinder effective RF signal capture in dynamic or low-
signal environments. Although L-type MN is widely adopted for its simple structure, low insertion loss and high
efficiency at a single frequency [18-21], it remains impractical at high frequencies [10]. Furthermore, only a few
RF-EH systems have employed multi-stage rectifiers, which are essential for boosting output voltage to meet
the demands of low-power devices [20-21]. Consequently, the review recommends that advanced RF-EH
systems incorporate high-gain antennas, optimised MNs using distributed elements and multi-stage rectifiers to
achieve sufficient output voltage and improve PCE.
This research proposes the design and optimisation of a 2.4 GHz RF-EH system for low-power IoT applications.
The system integrates a high-gain 4x4 MPA array for efficient RF signal capture and a 5-stage CW rectifier with
a single-stub MN. The multi-stage topology is employed to maximise the DC output voltage and PCE, thereby
improving the system’s reliability to meet the energy demands of low-power IoT devices. The single-stub MN
was also utilised to minimise signal reflection and ensure maximum PT between the antenna and the rectifier.
The remainder of this paper is organised as follows. The materials and methods used are presented in Section 2.
Followed by results and discussion in Section 3 and the conclusion in Section 4.
MATERIALS AND METHODS USED
Receiver Antenna Design
Single-Element Patch Antenna
Figure 2 shows the design of a single-element (1x1) patch antenna, which served as a template for developing
the 4x4 MPA array. The antenna was mounted on Rogers RT5880 substrate (dielectric constant = 2.2, loss
tangent = 0.0009, height = 1.588 mm). Both the patch and the ground plane were made of annealed copper with
a thickness of 0.035 mm, and the total area of the antenna was 90 × 100 mm². Rogers RT5880 substrate was
selected for its low dielectric constant and loss tangent, which provides better IM, reduced signal loss and a more
precise radiation pattern, making it suitable for high-frequency applications. Annealed copper was chosen for its
excellent conductivity and minimal resistive losses.
A 50 Ω microstrip feed line (length Lf = 22.84 mm, width Wf = 4.89 mm) was used to feed the antenna with
optimal power for normal operation, while a full ground plane ensured improved directivity and reduced back-
lobe radiation.
Both the ground plane and substrate shared the same dimensions (Lg = Ls = 90 mm, Wg = Ws =
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 344
100 mm). A recessed distance of length Lif = 9.16 mm and a feed gap of width Wif = 9 mm were cut into the
patch to ensure that the antenna resonates at 2.4 GHz with excellent IM. The antenna dimensions were derived
from the standard design equations provided below:
Patch width:
r
o
f ε + 1
c2
Wp =
2
(1)
where, Wp = patch width
c = speed of light in a vacuum
fo = resonant frequency
εr = dielectric constant of the substrate
Patch length:
rr
reff
ε + 1 ε 1
ε = +
2
h
2 1 + 12
Wp
(2)
reff
reff
Wp
+ 0.264
ε + 0.3
h
ΔLp = 0.412h
Wp
ε 0.258
+ 0.8
h
(3)
reffo
eff
L
2f
p =
ε
c
(4)
eff
Lp Lp 2ΔLp
(5)
where, εreff = effective dielectric constant
h = substrate height
Lp = actual length of patch
ΔLp = extension of patch length
Lpeff = effective patch length
Substrate length and width:
Ls = Lp + 6h
(6)
Ws = Wp + 6h
(7)
where, Ls = Length of the substrate
Ws = Width of the substrate
Feed line length and width:
gLf = 0.25λ
(8)
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 345
reff
o
g
λ
λ =
ε
(9)
o
o
c
f
=λ
(10)
r
o
ε +1.41
Z
87
t
7.48h
Wf = 1.25t
e
(11)
where, Lf = length of the feed line
λg = guided wavelength
λo = wavelength of the resonating frequency
Wf = width of the feed line
Zo = impedance of the feed line
tt = trace thickness
Recessed distance:
-1
o
in
Z
Lp
Lif = cos
πZ
(12)
where, Lif = recessed distance
Lp = length of the patch
Zo = characteristic impedance of the microstrip feed line
Zin = input impedance of the patch edge
Figure 2. Design of Single-Element (1x1) Patch Antenna
Proposed 4x4 Microstrip Patch Antenna Array
Figure 3 shows the design of the proposed 4x4 MPA array, which was created by arranging four 1x4 linear arrays
in parallel and feeding them from a single feed point. The array was mounted on a full ground plane of the same
area as the dielectric substrate. A 50 Ω microstrip feed line was first created along one of the edges of the
substrate extending towards the centre. The feed line dimensions were length Lmf = 200 mm and width Wmf =
4.89 mm and served as the main feed to the identical patch elements.
A T-junction (tj) feed line was created at
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 346
the end of the main feed line to divide the power coming from the main feed line equally among four sub-arrays
consisting of four patch elements each. This T-junction is labelled (tj_4). This tj_4 feed line was connected to
tj_3 via a pair of quarter-wave transformer (QwT) lines (Q_4) and the connector (C_4) at both ends. Tj_3 further
splits and conveys power equally to two pairs of patch elements within each sub-array through Q_3 and C_3. At
the ends of C_3, the power is again equally divided using Tj_2 through Q_2 and C_2, leading to Tj_1. Finally,
at Tj_1, power is distributed equally to each patch element through dedicated patch feed lines. The physical
dimensions of the 4x4 MPA array are provided in Appendix A.
The addition of more patch elements were adopted to enhance directivity and gain by receiving more power in
a specified direction. Moreover, careful arrangement of identical patch elements, with controlled amplitudes and
phases, improves radiation efficiency and shapes the overall radiation pattern.
Figure 3. Design of 4x4 Microstrip Patch Antenna Array
Matching Network Design
The design of a MN is to ensure maximum PT between the antenna and the rectifier circuit. This is achieved by
transforming the complex impedance of the rectifier to match the characteristic impedance of the antenna, which
is usually 50 Ω. A 5-stage CW rectifier was first designed, and its input impedance was measured by plotting
Zin against frequency. At 2.4 GHz, the rectifier input impedance was obtained as (3.815 – j37.583) Ω, which
served as a reference for designing the MN.
Owing to the degraded performance of lumped elements at high frequencies, a single-stub MN based on
microstrip TLs was adopted for a more practical and high-performance design. An ideal TL model was designed
using the Smith Chart Utility (SCU) in ADS to transform the rectifier impedance to 50 Ω. The impedances and
electrical lengths of the ideal TLs were obtained as (50 Ω, 13.09°) for TL1 and (50 Ω, 24.56°) for TL2, as shown
in Figure 4. These values serve as initial design parameters for further implementation using microstrip lines and
subsequent optimisation. TL1 and TL2 represent a short-circuit stub (SCS) and a series microstrip line (SML),
respectively.
For practical implementation, the ideal TLs were converted to real microstrip lines. This process began by
defining the substrate on which the microstrip lines would be etched.
The MSub component was selected from
the component palette, and its parameters were entered. Rogers RT5880 was selected as the substrate, consistent
with the antenna substrate, to maintain impedance continuity and minimise parasitic effects. The LineCalc option
in the Tools menu was used to convert the ideal TL parameters into their corresponding physical dimensions
(length and width) based on the defined substrate properties. The resulting single-stub MN consisted of the
following components: TL3, TL5, MTEE and MLSC.
Simulation of the initial layout produced an S11 value of -1.866 dB at 2.4 GHz, indicating sub-optimal IM. To
enhance performance, optimisation of the TL dimensions was performed. The widths of all the TLs were fixed
at 4.89 mm to match the antenna feed line, while the length of TL5 was set to 2 mm. Prior to optimisation, the
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 347
stub length (Lc) and SML length (Ld) were obtained as 3.31 mm and 6.21 mm, respectively. An optimisation
goal of S11
< -60 dB at 2.4 GHz was defined, with Lc and Ld varied between 1 mm and 20 mm. The Quasi-
Newton (QN) algorithm was employed with 100 iterations and a target error of 0.00001. Following optimisation,
the MN demonstrated significantly improved matching characteristics. Figure 5 presents the single-stub MN for
the 5-stage CW rectifier after optimisation.
Figure 4. Ideal Transmission Line Model of Single-Stub Matching Network
Figure 5. Single-Stub Matching Network for 5-Stage CW Rectifier After Optimisation
Rectifier Circuit Design
The
rectifier converts harvested RF energy from the antenna into usable DC output. In this research, a 5-stage
rectifier based on the Cockcroft-Walton Voltage Multiplier (CW-VM) topology was implemented to achieve
high PCE and appreciable DC output voltage under low input power conditions. Each stage of the CW rectifier
consists of two Schottky diodes and a pair of capacitors. Zero-bias HSMS-2850 Schottky diodes were chosen
for their low turn-on voltage (150-250 mV) and fast switching performance at high frequencies, making them
suitable for RF-EH applications. The utilisation of Schottky diodes ensures minimal voltage loss during
rectification, resulting in improved PCE, while the capacitors in the circuit store and transfer charge between
stages, enabling voltage multiplication.
The circuit model of the zero-bias Schottky diode is illustrated in Figure 6. It consists of a 0.7 nH series inductor
and a 0.25 pF shunt capacitor to account for parasitic effects arising from the diode’s bond wires, package leads
and stray capacitance. Model parameters (DIODEM1), derived from the manufacturer’s data sheet, ensured
accuracy during Harmonic Balance (HB)
simulations [21]. These parameters are displayed in Table 1.
The schematic diagram of the 5-stage CW rectifier is presented in Figure 7. A custom diode model symbol was
created from the HSMS-2850 model and reused across subsequent stages. A separate schematic was then
designed to determine the optimal load resistance and input RF power that maximises the rectifier’s PCE, as
well as the input impedance associated with these values. This new schematic implemented a 5-stage CW-VM
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 348
topology having two of the custom-defined diode models and a pair of 150 pF series capacitors for each stage
with a 0.047 pF shunt capacitor at the output to act as a low-pass filter.
Figure 6. Circuit Model of Zero-Bias Schottky Diode
Table 1.
Parameters of Zero-Bias HSMS-2850 Schottky Diode
Parameter
Value
Unit
Minimum forward voltage
0.15
V
Saturation current
3
µA
Saturation maximum reverse leakage current
150
µA
Breakdown voltage
3.75
V
Parasitic series resistance
25
Ω
Junction capacitance
0.18
pF
Detection sensitivity
50
mV/µW
Figure 7. Schematic Diagram of Proposed 5-Stage CW Rectifier
The rectifier was driven by a single-tone RF source (P_1Tone) at 2.4 GHz with a characteristic impedance of 50
Ω. Input power was controlled by the variable pin (in dBm), allowing for flexible power sweep analysis. An HB
analyser was used to model nonlinear diode behaviour, with harmonics up to the fifth order (Order[1] = 5)
considered to capture rectification dynamics and harmonic content.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 349
Initial simulations were performed without the MN, and the input impedance of the rectifier was obtained as
(3.815 – j37.583) Ω.
To match the rectifier’s input impedance with the 50 Ω source, a single-stub MN was
designed using the Smith Chart Tool in ADS. S-parameter simulation results confirmed that the MN successfully
transformed the source impedance to match the rectifier’s input impedance at 2.4 GHz. With the MN
(DA_SmithChartMatch2) integrated, the rectifier demonstrated significant improvements in both output DC
voltage and PCE. Optimal performance was observed with a 300 kΩ load resistance, where cascaded stages
combined with precise IM substantially enhanced rectification efficiency.
RESULTS AND DISCUSSION
Antenna Performance
The return loss represents the portion of incident power reflected due to impedance mismatch between the
antenna and its feedline. For effective operation, values below -10 dB are required. As shown in Figure 8,
both
antennas achieve return loss values below -25 dB at 2.4 GHz: -33.204 dB for the 1x1 patch and -25.56 dB for
the 4x4 array. The slightly less negative value of the 4x4 array is attributed to the complexity of its feed structure
and mutual coupling effects of adjacent array elements [7]. The measured bandwidths were 41.4 MHz (2.3793
– 2.4207 GHz) for the 1x1 patch and 33.5 MHz (2.3843 – 2.4178 GHz) for the 4x4 array, both sufficient for
ISM-band operation. The Voltage Signal Wave Ratio (VSWR) is also an important parameter that specifies how
well the antenna is matched with its feedline. A VSWR between 1 and 2 indicate acceptable matching [11]. As
shown in Figure 9, the 1x1 patch and the 4x4 array achieved 1.0447 and 1.1113, respectively, both close to unity,
confirming minimal reflection losses.
Figure 10 shows that the 1x1 patch achieved a gain of 7.122 dBi, which increased to 19.29 dBi with the 4x4
array due to its large aperture and constructive interference of EM waves from multiple, closely spaced elements.
Directivity also improved, from 8.081 dBi for the 1x1 patch to 20.01 dBi for the 4x4 array as seen in Figure 11.
The measured radiation efficiencies were -0.9586 dB (80.19 %) for the 1x1 patch and -0.7218 dB (84.69 %) for
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 350
the 4x4 array, while total efficiencies were -0.9712 dB (79.96 %) and -0.7346 dB (84.44 %), respectively.
Despite coupling losses, the 4x4 array achieved over 84 % of total and radiation efficiencies, meeting RF-EH
requirements. Efficiency in percentage was determined using equation (13):
0.1 x (Rad or Tot Efficiency) dB
Efficiency (%) = 10
(13)
where, Efficiency (%) = Efficiency in percentage
(Rad or Tot Efficiency) dB = Radiation or Total Efficiency in dB
The E-plane radiation pattern (φ = 0°) shows the elevation-plane behaviour. As shown in Figure 12, the 1x1
patch exhibited a main lobe magnitude of 7.14 dBi at 2°, with a 77.2° beamwidth and a -19.2 dB side lobe. The
4x4 array produced a narrower 19.4
° beamwidth, a 19.3 dBi main lobe at 0° and a -13.8 dB side lobe,
corresponding to only 4.2 % power loss into undesired directions. The H-plane radiation pattern (
φ = 90°)
exhibits similar improvements as seen in Figure 13. The 1x1 patch achieved 7.12 dBi with a 70.9
° beamwidth
and -20.8 dB side lobe, while the 4x4 array reached 19.3 dBi with a 15.9
° beamwidth and -12.9 dB side lobe.
Overall, the 4x4 array demonstrated superior gain, directivity, radiation and total efficiencies, and highly focused
E and H plane patterns, confirming its suitability for efficient RF-EH.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 351
Figure 13. H-Plane Radiation Pattern
Matching Network Performance
Figure 14 illustrates the return losses of the single-stub MN for the 5-stage CW rectifier before and after
optimisation. Using the QN algorithm, the physical dimensions of the SCS stub and SML were adjusted for
optimal IM at 2.4 GHz. The initial return loss was -1.866 dB, indicating poor matching. After optimisation, it
improved significantly, to -62.858 dB. The QN algorithm gave a final value of -62.8599 dB within only five
iterations, demonstrating higher accuracy and efficiency. The optimised lengths were Lc = 3.37688 mm and Ld
= 4.236 mm for SCS and SML, respectively, as shown in Figure 15. These refinements fine-tuned the MN for
maximum PT to the rectifier. These results confirm the precision advantage of QN over alternative ADS
optimisation methods, which explains why it is commonly used for single-frequency S11 parameter
optimisation.
Figure 16 presents the real and imaginary components of the 5-stage CW rectifier before and after matching. At
2.4 GHz, the input impedance was (3.815 – j37.583) Ω, indicating severe mismatch. After integrating the single-
stub MN, the impedance transformed to (50.221 – j0.052) Ω, where the real part closely approached
50 Ω and
the imaginary part nearly cancelled out, signifying successful IT. Figure 17 illustrates the return loss, which
improved from -0.847 dB (unmatched) to -52.907 dB (matched). The results demonstrate a near-perfect IM, with
minimal power loss between the rectifier and the 50
Ω RF source. The single-stub MN therefore proves effective
and reliable for integration into the cascaded 5-stage rectifier circuit.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 352
Figure 16. Input Impedance of 5-Stage CW Rectifier
(b) After Optimisation (a) Before Optimisation
(b) After Matching
(a) Before Matching
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 353
Rectifier Performance
Figure 18 shows the variation of input power (Pin) with output DC voltage for the proposed 5-stage CW rectifier.
The circuit exhibits a steady increase in output voltage across the measured Pin range, beginning with 0.456 V
at -20 dBm, which corresponds to extremely weak input RF signals typically observed in ambient IoT
environments. At -10 dBm, the output DC voltage rises to 1.670 V, while at 0 dBm it achieves 4.144 V,
demonstrating its capability to harvest sufficient energy for low-power sensing applications. A further increase
to 10 dBm results in 9.816 V, and the maximum recorded output is 16.424 V at 17.5 dBm.
Figure 19 presents the rectifier’s PCE as a function of Pin. The efficiency improves gradually, starting from
0.069 % at -20 dBm and 0.929 % at -10 dBm, which corresponds to ambient RF power levels typically available
in low-power scenarios. At 0 dBm, the PCE rises to 5.725 %, and at 10 dBm, the rectifier achieves 32.116 %,
before reaching a maximum efficiency of 89.913 % at 17.5 dBm.
The results obtained are significant for two reasons. First, they confirm that the rectifier operates optimally at
very low Pin levels, which is a critical requirement for practical IoT deployments where available RF energy is
often below -10 dBm. Second, they reveal that the rectifier is able to scale its output effectively under stronger
excitations, until it reaches a saturation point beyond 17.5 dBm, where additional increases in Pin result in
negligible voltage improvement. The observed saturation around 16.4 V highlights both the efficiency and the
practical operation limits of the design.
The proposed rectifier contributes significantly to this performance enhancement by boosting the output voltage,
as previously discussed. While the inclusion of multiple diodes and capacitors inherently introduces conduction
losses and slightly reduces efficiency at very low Pin levels, the design, however, ensures that the harvested
energy remains usable. This trade-off between low PCE and high DC output voltage demonstrates the design’s
practical balance. Importantly, the ability to sustain measurable PCE even under weak input conditions confirms
the circuit’s suitability for RF-EH in environments where available RF power is scarce but still sufficient for
intermittent or low-duty-cycle IoT applications.
Moreover, from a practical viewpoint, the proposed design can be readily fabricated using commercially
available low-cost substrates such as Rogers or standard FR-4, depending on the desired balance between
performance and cost. Rogers RT 5880 employed in this study offers low dielectric losses at high frequencies.
However, FR-4 provides a more economical solution for large-scale IoT deployments, albeit with slightly
reduced performance. The circuit relies on off-the-shelf Schottky diodes and surface-mount capacitors, which
ensures ease of prototyping and scalability to industrial production without requiring complex CMOS
fabrication.
(b) After Matching
(a) Before Matching
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 354
Figure 18. Pin Versus DC Output Voltage of Proposed 5-Stage CW Rectifier
Figure 19.
Pin Versus PCE of Proposed 5-Stage CW Rectifier
Comparative Analysis
The section provides a comparative analysis of the proposed RF-EH system against related works discussed in
the literature review. The analysis is divided into two parts: the first assesses the performance of the receiver
antenna, and the second evaluates the rectifier circuit. Table 2 compares the proposed 4x4 MPA array with the
previously reported antenna designs, while Table 3 summarises the comparative performance of the proposed 5-
stage rectifier.
Table 2 Comparison of Proposed 4x4 MPA Array with Existing Antenna Designs
SN
Reference
Frequency
(GHz)
Substrate
Array
Size
Return
Loss (dB)
Gain
(dBi)
Rad. Eff.
(%)
Tot. Eff.
(%)
1.
[9]
2.4, 5.8
FR-4
1x1
-14.104, -
27.515
0.5102,
1.175
27, 38
25, 38
2.
[8]
2.4, 5.2
FR-4
1x2
-20.6, -16
1.6, 3.95
81, 80
69, 67
3.
[3]
1.83, 2, 2.45
Arlon
1x1
< -10
3.5, 5, 5
NA
NA
4.
[14]
1.8
FR-4
2x2
-25
9.2
NA
NA
5.
[15]
2.4
FR-4
1x2, 2x2,
2x3, 2x4
< -10
7.2, 11,
10.4, 10.3
NA
NA
6.
[7]
2.45
Rogers
RT5880
3-element
-32.886
11.24
89.45
89.41
7.
[12]
2.3
FR-4
1x1, 2x2,
4x4
-15.4, -
17.6, -
17.10
3.05, 7.65,
13.05
NA
NA
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 355
8.
[16]
2.4
Rogers
RT5880
7-element
< -50
15.8
96
94
9.
[17]
2.4
Rogers
RO4003
2x4
-26.2
17.5
NA
NA
10.
This work
2.4
Rogers
RT5880
4x4
-25.56
19.29
84.69
84.44
Table 3 Comparison of Proposed Five-Stage Rectifier with Existing Rectifier Designs
S
N
Author
(Year)
Frequency
(GHz)
Diode
Type
Input
Power
(dBm)
Number
of Stages
Load
Resistor
(k
)
Maximum
PCE (%)
Output
Voltage
(V)
1.
[18]
2.4
HSMS-
2850
10
1
2
75
2.9
2.
[6]
2.45, 5.5
SMS7630
-061
-10
1
5
47, 27
0.51, 0.36
3.
[19]
2.45
HSMS-
2850
19
1
20
77.57
1.092
4.
[20]
2.45
HSMS-
2850
0
2
470
64
5.2
5.
[21]
2.45
HSMS-
285C
20
7
1000
75
12
6.
This work
2.4
HSMS-
2850
17.5
5
300
89.913
16.424
CONCLUSION
A 2.4 GHz RF-EH system was designed and optimised to provide a sustainable power solution for low-power
IoT applications. This work focused on three key components: a high-gain 4x4 MPA array, an optimised single-
stub MN and a cascaded 5-stage CW rectifier for efficient RF-to-DC conversion. The proposed 4x4 antenna
array achieved a gain of 19.29 dBi and directivity of 20.01 dBi with radiation and total efficiencies exceeding
84 %, enabling effective RF signal capture. The single-stub MN achieved precise IM by transforming the
rectifier’s input impedance from (3.815 – j37.583) Ω to approximately 50 Ω, resulting in a return loss of -62.858
dB. This ensured minimal reflection and maximum PT between the antenna and rectifier. The cascaded 5-stage
rectifier produced usable DC output voltages ranging from 0.456 V at -20 dBm to 4.144 V at 0 dBm,
demonstrating strong suitability for IoT devices operating under limited RF availability. At higher input powers,
the rectifier achieved a peak output voltage of 16.424 V with a PCE of 89.913 % at 17.5 dBm. Ultimately, the
proposed RF-EH system provides a sustainable, cost-effective and eco-friendly power solution for large-scale
IoT and Wireless Sensor Network (WSN) deployments, reducing the reliance on traditional batteries while
ensuring efficient RF-to-DC power conversion.
To further enhance system robustness, future work could explore
hybrid energy harvesting approaches, where RF-EH is complemented with solar, thermal or vibration modalities.
Such multi-source architectures would provide more reliable and sustainable power delivery, enabling support
for more demanding IoT applications under diverse environmental conditions.
REFERENCE
1. Moloudian, G., Hosseinifard, M., Kumar, S., Simorangkir, R. B. V. B., Buckley, J. L., Song, C., Fantoni,
G and O’Flynn, B. (2024), “RF Energy Harvesting Techniques for Battery-less Wireless Sensing,
Industry 4.0 and Internet of Things: A Review”, IEEE Sensors Journal, Vol. 24, No. 5, pp. 1 – 14.
2. Kanboz, B. and Palandöken, M. (2023), “Microstrip Patch Antenna Array Design for RF Energy
Harvesting Applications”, European Journal of Science and Technology, Vol. 3, No. 49, pp. 34 – 37.
3. Zeng, M., Li, Z., Ding, Y. and Lin, X. (2022), “A Wide‐Band Antenna with Dual Open Rings Resonator
for RF Energy Harvesting”, Microwave and Optical Technology Letters, Vol. 64, No. 11, pp. 2019 –
2023.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 356
4. Agrahari, R., Singh, S., Samantaray, D., Kumar, B., Bhattacharyya, S., Mahto, M. and Jain, P. K. (2023),
“Triple‐Band Meta-surface Absorber for RF Energy Harvesting Applications”, Microwave and Optical
Technology Letters, Vol. 65, No. 8, pp. 2252 – 2261.
5. Lee, Y. C., Ramiah, H., Choo, A., Churchill, K. K. P., Lai, N. S., Lim, C. C., Chen, Y., Mak, P. and
Martins, R. P. (2023), “High-Performance Multiband Ambient RF Energy Harvesting Front-End System
for Sustainable IoT Applications – A Review”, IEEE Access, Vol. 11, No. 1, pp. 11143 – 11164.
6. Mehta, P. and Nella, A. (2024), “Dual-Band Low-Power RF-to-DC Signal Converter Circuits for Energy
Harvesting”, AIP Advances, Vol. 14, No. 7, pp. 1 – 13.
7. Ansah, J. N. and Annan, J. K. (2024), “A Novel Triangular Configuration of a Rectangular Patch Antenna
Array For 2.45 GHz RF Energy Harvesting”, International Research Journal of Engineering and
Technology, Vol. 11, No. 11, pp. 420 – 427.
8. El-Issawi, M. L., Konditi, D. B. O. and Usman, A. D. (2024), “Design of an Enhance Dual-Band
Microstrip Patch Antenna with Defected Ground Structures for WLAN and WiMAX”, Indonesian
Journal of Electrical Engineering and Computer Science, Vol. 35, No. 1, pp. 165 – 174.
9. Küçükcan, S. and Kaya, A. (2022), “Dual-Band Microstrip Patch Antenna Design for Wi-Fi
Applications”, European Journal of Science and Technology, Vol. 2, No. 34, pp. 661 – 664.
10. Luo, Y., Pu, L., Wang, G. and Zhao, Y. (2019), “RF Energy Harvesting Wireless Communications: RF
Environment, Device Hardware and Practical Issues”, Sensors, Vol. 19, No. 13, 3010pp.
11. Huang, Y. and Boyle, K. (2008), Antennas: From Theory to Practice, John Wiley and Sons Ltd,
Chichester, United Kingdom, (ed.), 2
nd
Edition, pp. 1 – 363.
12. Wijayanti, A., Putri, T. L., Puspitorini, O. and Siswandari, N. A. (2024), “Design of Microstrip Planar
Array Antenna for Wireless Sensor Device Charging at Frequency 2300 MHz”, Proceedings of the
International Conference on Applied Science and Technology on Engineering Science 2023 (iCAST-ES
2023), Dordrecht, Netherlands, pp. 493–505.
13. Kuşin, Y. E. and Palandöken, M. (2023), “Machine Learning Based B-Shaped Monopole Antenna for
RF Energy Harvesting Applications”, International Journal of Advanced Natural Sciences and
Engineering Researches, Vol. 7, No. 2, pp. 1 – 8.
14. Sharma, T. and Saini, G. (2016), “Microstrip Antenna Array for RF Energy Harvesting System”,
International Journal of Advanced Information Science and Technology, Vol. 45, No. 45, pp. 145 – 149.
15. Jayanthy, T., Harini, J., Haseena, A. and Neha, C. L. (2024), “Design and Analysis of Triangular Patch
Antenna Array”, International Journal of Engineering Applied Sciences and Technology, Vol. 8, No. 12,
pp. 123 – 128.
16. Kabeel, A. A., Hussein, A. H., ElMaghrabi, A. E. and Elabd, R. H. (2023), “Design of a Circular
Concentric Microstrip Patch Antenna Array for Wi-Fi Band Energy Harvesting”, Journal of Engineering
Research, Vol. 7, No. 5, pp. 156 – 159.
17. Nnamdi, U., Omijeh, B. and Asianuba, I. (2023), “Design and Optimisation of a 2.4 GHz Antenna Array
for Energy Harvesting”, European Journal of Theoretical and Applied Sciences, Vol. 1, No. 6, pp. 676 –
683.
18. Subbyal, H., Ali, W. and Liguo, S. (2022), “Compact Antenna Integrated with a Greinacher Voltage
Multiplier for Ambient Energy Harvesting Applications”, International Journal of RF and Microwave
Computer-Aided Engineering, Vol. 32, No. 12, pp. 1 – 10.
19. Pramono, S., Shidiq, D. D., Ibrahim, M. H., Adriyanto, F. and Hikmaturokhman, A. (2021), “RF Energy
Harvesting using a Compact Rectenna with an Antenna Array at 2.45 GHz for IoT Applications”, Journal
of Electrical Engineering, Vol. 72, No. 3, pp. 159 – 167.
20. Khan, N. U., Ullah, S., Khan, F. U. and Merla, A. (2024), “Development of 2400-2450 MHz Frequency
Band RF Energy Harvesting System for Low-Power Device Operation”, Sensors, Vol. 24, No. 10, pp. 1
– 13.
21. Omara, F. A., Ali, W. A. E., Eltrass, A. S. and Abbasy, N. H. (2024), “Design of a 2.45 GHz Rectenna
Based on High-Gain 2-Element Array Antenna and 7-stage Voltage-Doubler Rectifier for Energy
Harvesting Applications”, Physica Scripta, Vol. 99, No. 12, pp. 1 – 20.
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN APPLIED SCIENCE (IJRIAS)
ISSN No. 2454-6194 | DOI: 10.51584/IJRIAS |Volume X Issue IX September 2025
www.rsisinternational.org
Page 357
APPENDIX A
Table A1. Dimensions of the 4x4 MPA Array
Parameter
Value (mm)
Parameter
Value (mm)
Ws
400
Ls
360
Wg
400
Lg
360
Wp
51
Lp
42.35
Wmf
4.89
Lmf
200
Wif
8.90
Lif
10
Wtj_1
30
Ltj_1
1.42
Wq_1
20
Lq_1
2.81
Wc_1
20
Lc_1
4.89
Wpf
4.89
Lpf
20
Wtj_2
1.42
Ltj_2
20
Wq_2
2.81
Lq_2
15
Wc_2
4.89
Lc_2
15
Wtj_3
130
Ltj_3
1.42
Wq_3
20
Lq_3
2.81
Wca_3
15
Lca_3
4.89
Wcb_3
4.89
Lcb_3
26
Wtj_4
1.42
Ltj_4
85
Wq_4
2.81
Lq_4
20
Wc_4
4.89
Lc_4
20
d
110
