Periodic Reporting for period 3 - APOLLO (Advanced Signal Processing Technologies for Wireless Powered Communications)
Période du rapport: 2022-07-01 au 2023-12-31
Wireless powered communication (WPC) technology has been recently proposed as a means to realize energy-sustainable networks via the transmission of radio frequency (RF) signals by dedicated sources for both information and energy transfer purposes. As such, it is regarded as an enabler for the Internet-of-Things (IoT) paradigm and an integral component of future 5th Generation (5G) releases and beyond (B5G). Consequently, WPC supports the provision of advanced services to individuals as well as to a broad array of diverse vertical industries, thereby benefiting the economy and the society.
WPC represents a paradigm shift that requires a redesign of wireless communication networks. APOLLO tackles all the issues associated with the realization of the WPC vision. Specifically, its objectives include:
• the development of realistic yet tractable models for RF energy harvesting, the establishment of fundamental performance limits, the characterization of the rate-energy trade-off, and the design of waveforms.
• the development and performance evaluation (both analytically and numerically) at link and system level of physical (PHY) layer, medium access control (MAC) layer, and resource allocation schemes and protocols.
• the hardware implementation and experimental validation of promising techniques.
Furthermore, APOLLO focuses on the main applications of the WPC paradigm mentioned above, namely, machine-to-machine (M2M) communications and (B)5G networks. Small cells, mobile edge caching, cellular and cell-free massive multiple-input multiple-output (MIMO), millimetre-wave (mmWave) communications, and intelligent reflecting surfaces (IRS) are representative examples of the considered technologies.
The APOLLO brings together a team of experts on complementary disciplines such as information theory, signal processing, stochastic geometry, probability theory, and optimization theory; targets a holistic study of all WPC research areas; investigates the integration of the WPC paradigm with innovative M2M and (B)5G technologies; and considers multi-user/multi-cell/multi-antenna communication setups, thus deviating from the rather simplistic early works on the subject. In this way, the project aspires to: 1) push the boundaries of WPC well beyond its current capabilities; 2) shed light on the performance, design, testing, and development of contemporary and future WPC systems; and 3) facilitate the transition of this radically new approach in wireless communications from its conceptual phase to practical implementation.
WPC represents a paradigm shift that requires a redesign of wireless communication networks. APOLLO tackles all the issues associated with the realization of the WPC vision. Specifically, its objectives include:
• the development of realistic yet tractable models for RF energy harvesting, the establishment of fundamental performance limits, the characterization of the rate-energy trade-off, and the design of waveforms.
• the development and performance evaluation (both analytically and numerically) at link and system level of physical (PHY) layer, medium access control (MAC) layer, and resource allocation schemes and protocols.
• the hardware implementation and experimental validation of promising techniques.
Furthermore, APOLLO focuses on the main applications of the WPC paradigm mentioned above, namely, machine-to-machine (M2M) communications and (B)5G networks. Small cells, mobile edge caching, cellular and cell-free massive multiple-input multiple-output (MIMO), millimetre-wave (mmWave) communications, and intelligent reflecting surfaces (IRS) are representative examples of the considered technologies.
The APOLLO brings together a team of experts on complementary disciplines such as information theory, signal processing, stochastic geometry, probability theory, and optimization theory; targets a holistic study of all WPC research areas; investigates the integration of the WPC paradigm with innovative M2M and (B)5G technologies; and considers multi-user/multi-cell/multi-antenna communication setups, thus deviating from the rather simplistic early works on the subject. In this way, the project aspires to: 1) push the boundaries of WPC well beyond its current capabilities; 2) shed light on the performance, design, testing, and development of contemporary and future WPC systems; and 3) facilitate the transition of this radically new approach in wireless communications from its conceptual phase to practical implementation.
The work performed so far within the APOLLO project has advanced the research on WPC and allowed for a deeper understanding of such systems. This can be seen through the 64 publications in prestigious international journals and conference proceedings, and through the participation of the PI in several keynotes/tutorials at various conferences (e.g. IEEE GLOBECOM, IEEE PIMRC, IEEE VTC etc) and training schools on wireless communications.
The current achievements of the APOLLO project can be summarized as follows:
1. The research in APOLLO has tackled many fundamental theoretical studies of WPC. Among the contributions, new mathematical models for the rectification process that performs energy harvesting have been studied. Moreover, and based on tools from information theory, the information-energy capacity region for various types of channels has been established. Finally, waveforms and modulation schemes that enhance the simultaneous wireless information and power transfer (SWIPT) performance have been proposed.
2. The integration of WPC in 5G/B5G has been considered. The analysis, modeling, and design of WPC-enabled modern technologies such as massive MIMO, mmWave networks, edge computing, and IRSs using tools from optimization theory, stochastic geometry, and probability theory have led to innovative communication schemes.
3. The combination of WPC with machine- type communications has been investigated. More specifically, WPC-enabled LoRa networks, sensor networks, and mobile cellular networks with ambient RF energy harvesting have been considered. As an example, a framework that allows nodes to optimally access the channel in a passive IoT network, where nodes harvest energy to communicate, has been proposed. Moreover, the co-design of WPC with federated learning as well as the energy multicasting operation of magnetic resonance have been studied.
4. A novel frequency domain SWIPT waveform design that allows to enhance the WPT performance without impacting the information transfer performance has been implemented in software-defined radio (SDR). On the other hand, a graphene-based IRS that allows for active and dynamic control of THz waves, suitable for WPC, has been investigated.
5. Modern communication setups representing key-enabling technologies for future networks have been also investigated in the project, although not explicitly considering WPC. However, these works have allowed to get more insight onto such systems which helped in facilitating other SWIPT-related works.
The current achievements of the APOLLO project can be summarized as follows:
1. The research in APOLLO has tackled many fundamental theoretical studies of WPC. Among the contributions, new mathematical models for the rectification process that performs energy harvesting have been studied. Moreover, and based on tools from information theory, the information-energy capacity region for various types of channels has been established. Finally, waveforms and modulation schemes that enhance the simultaneous wireless information and power transfer (SWIPT) performance have been proposed.
2. The integration of WPC in 5G/B5G has been considered. The analysis, modeling, and design of WPC-enabled modern technologies such as massive MIMO, mmWave networks, edge computing, and IRSs using tools from optimization theory, stochastic geometry, and probability theory have led to innovative communication schemes.
3. The combination of WPC with machine- type communications has been investigated. More specifically, WPC-enabled LoRa networks, sensor networks, and mobile cellular networks with ambient RF energy harvesting have been considered. As an example, a framework that allows nodes to optimally access the channel in a passive IoT network, where nodes harvest energy to communicate, has been proposed. Moreover, the co-design of WPC with federated learning as well as the energy multicasting operation of magnetic resonance have been studied.
4. A novel frequency domain SWIPT waveform design that allows to enhance the WPT performance without impacting the information transfer performance has been implemented in software-defined radio (SDR). On the other hand, a graphene-based IRS that allows for active and dynamic control of THz waves, suitable for WPC, has been investigated.
5. Modern communication setups representing key-enabling technologies for future networks have been also investigated in the project, although not explicitly considering WPC. However, these works have allowed to get more insight onto such systems which helped in facilitating other SWIPT-related works.
The work performed so far in APOLLO has significantly advanced the state-of-the-art in all areas of WPC. Specifically:
• Rectification modeling and fundamental performance limits: The impact of energy harvesting distribution uncertainty, low-pass resistor-capacitor filter at the receiver, and non-linear high-power amplifier at the transmitter on the performance of wireless power transfer (WPT) has been studied.
• Waveform design: Novel SWIPT waveforms have been proposed, namely, a tone-index multisine waveform that achieves SWIPT with a passive receiver and a circular QAM modulation for SWIPT that enhances both the error rate and the energy rate in comparison to conventional modulation schemes. Furthermore, the design of chaotic waveforms for SWIPT has been investigated for the first time.
• Integration with 5G/B5G networks and M2M communication: The joint design of precoding and power splitting for SWIPT under RF exposure constraints has been studied for the first time. A novel low-complexity hybrid precoding design for SWIPT based on the stochastic gradient descent algorithm has been proposed. A framework for joint robust active/passive beamforming optimization in IRS-aided SWIPT systems under inter-system interference constraints, with or without security constraints, and imperfect channel state information (CSI) has been developed for the first time. A new random rotation coding scheme for IRS-aided communication that converts space diversity to time diversity, thus enhancing reliability without requiring CSI, has been derived. Also, a novel channel access mechanism for passive IoT networks has been described. In addition, the combination of SWIPT with federated learning in IoT setups has been studied and the trade-off between the number of communication rounds and the time required to harvest sufficient amounts of energy has been investigated for the first time.
• Hardware implementation and testbeds: A novel frequency-domain SWIPT waveform that enhances the energy rate with negligible deterioration of the information rate/symbol error rate while avoiding at the same time the use of receive filters for demultiplexing of the information and energy signals has been proposed and validated experimentally via an SDR-based testbed.
• Rectification modeling and fundamental performance limits: The impact of energy harvesting distribution uncertainty, low-pass resistor-capacitor filter at the receiver, and non-linear high-power amplifier at the transmitter on the performance of wireless power transfer (WPT) has been studied.
• Waveform design: Novel SWIPT waveforms have been proposed, namely, a tone-index multisine waveform that achieves SWIPT with a passive receiver and a circular QAM modulation for SWIPT that enhances both the error rate and the energy rate in comparison to conventional modulation schemes. Furthermore, the design of chaotic waveforms for SWIPT has been investigated for the first time.
• Integration with 5G/B5G networks and M2M communication: The joint design of precoding and power splitting for SWIPT under RF exposure constraints has been studied for the first time. A novel low-complexity hybrid precoding design for SWIPT based on the stochastic gradient descent algorithm has been proposed. A framework for joint robust active/passive beamforming optimization in IRS-aided SWIPT systems under inter-system interference constraints, with or without security constraints, and imperfect channel state information (CSI) has been developed for the first time. A new random rotation coding scheme for IRS-aided communication that converts space diversity to time diversity, thus enhancing reliability without requiring CSI, has been derived. Also, a novel channel access mechanism for passive IoT networks has been described. In addition, the combination of SWIPT with federated learning in IoT setups has been studied and the trade-off between the number of communication rounds and the time required to harvest sufficient amounts of energy has been investigated for the first time.
• Hardware implementation and testbeds: A novel frequency-domain SWIPT waveform that enhances the energy rate with negligible deterioration of the information rate/symbol error rate while avoiding at the same time the use of receive filters for demultiplexing of the information and energy signals has been proposed and validated experimentally via an SDR-based testbed.