Negligible Energy Consumption Research Articles

Ultrafast Hybrid Computing Systems Enabled by Memristor‐Based Quadratic Programming Circuits

AbstractImplementing algorithms purely on digital computing platforms dramatically halts the performance of conventional computing systems. Revolutionary computing systems with extreme energy efficiency and high accuracy are demanded to handle the growing computing tasks. Here, the research on hybrid analog–digital computing platforms enabled by memristor‐based optimization solvers for achieving ultrafast computations is presented. By utilizing tunable memristors as parameters to solve linear programming (LP) and quadratic programming (QP) problems, a real‐time control algorithm for micro air vehicles (MAVs) and a support vector machine (SVM) algorithm for cancer diagnosis are implemented. These experiments demonstrate over 2000x speed‐up compared to conventional digital platforms, with negligible energy consumption, using a memristor‐based system consisting of six memristors. These findings underscore the vast potential of memristor‐based optimization solvers not only in hybrid analog–digital computing platforms but also as a transformative solution for a wide range of modern computing challenges. This approach promises significant advancements in energy efficiency and ultrafast speed, positioning it as a leading contender for next‐generation computing paradigms.

PhotoSolver: A bidirectional photonic solver for systems of linear equations

Solving systems of linear equations is one of the most fundamental challenges in computing, with ubiquitous applications in science and engineering. Solving large-scale linear equations is particularly challenging, requiring tremendous computing expenses. However, conventional solvers in electronic digital computers face inevitable physical bottlenecks that hinder further advancements in electronic chips. Unlike the electronic architectures, optical computing harnesses the inherent advantages of lights, such as ultimately high speed, negligible energy consumption, and high parallelism, making optical architectures attractive sources for ultra-high speed analog computing. Especially, recent advances in integrated photonic circuits bring exciting new possibilities for high performance computing based on photonic chips. In this work, we propose a bidirectional PhotoSolver using the propagation of optical signals in different directions on a single integrated photonic chip to perform multiple operations in solving systems of linear equations, scalable to solve large-scale systems. Specifically, we propose an in situ approach to iteratively update solutions by performing the forward and backward propagations within a single photonic chip. Furthermore, we propose a partitioning approach to solve large-scale systems using arrays of photonic chips. Numerical experiments demonstrate these capabilities with representative systems of linear equations. By utilizing lights propagating in integrated photonic chips, our PhotoSolvers provide a promising computing architecture to solve large systems of linear equations, with potentials in wide computational applications.

Alkali-driven Donnan dialysis for efficient ammonia recovery from wastewater: Performance, mechanism and optimization

Recovery of ammonia from wastewater is of practical importance toward a sustainable society. To this end, concentration-driven Donnan dialysis (DD) is a promising recovery method especially for its simplicity and negligible energy consumption. Upon the traditional salt-driven DD, the conceptive alkali-driven DD significantly enhanced mass transfer and removal efficiency, yet their performance in terms of key process parameter were not deeply understood and further optimization remained unclear. Models for salt (NaCl) and alkali (NaOH)-driven DD processes were thus established and the ammonia transfer was theoretically investigated. The results showed that a fluctuated Donnan equilibrium constant was found responsible for the superior performance of alkali-driven DD (mass transfer accelerated by 90 % than control). Interestingly, there was an optimal ammonia recovery mode via adjusting ratios of alkali/ammonia and cation/ammonia, two key parameters governing the process efficiency, based on which the salt-alkali coupling-driven DD process was further proposed. It was verified that the coupling-driven DD process achieved the ammonia removal kinetic value of 0.1632 h−1 under insufficient alkali conditions (accounting for 50 % ammonia concentration in feed) with lowest CO2 emissions of 3.49 kg CO2/kg NH4+. The descending overall efficiency of ammonia removal, following alkali-driven (0.2943 h−1) > coupling-driven (0.1632 h−1) > salt-driven (0.1228 h−1), was ascribed to the variations in the electrochemical potential of ions involved in the respective processes. In addition, the cost of chemical consumption for a coupling-driven DD process (3.99 $/kg NH4+) was close to that of a salt-driven process, but much lower than that of an alkali-driven process (14.81 $/kg NH4+). These finding insight the low-carbon and efficient recovery of ammonia from wastewater.

Spontaneous and rapid electrostatic solvent nanofiltration based on a conductive layered membrane

An electrostatic solvent nanofiltration (ESN) method was developed for separating water from a variety of miscible solvents with high flux and negligible energy consumption by using an electrically conductive layered MXene membrane.

Internet-of-Things framework for scalable end-of-life condition monitoring in remanufacturing

The worldwide generation of waste electrical and electronic equipment is continuously growing, with electric vehicle batteries reaching their end-of-life having become a key concern for both the environment and human health in recent years. In this context, the proliferation of Internet of Things standards and data ecosystems is advancing the feasibility of data-driven condition monitoring and remanufacturing. This is particularly desirable for the end-of-life recovery of high-value equipment towards sustainable closed-loop production systems. Low-Power Wide-Area Networks, despite being relatively recent, are starting to be conceived as key-enabling technologies built upon the principles of long-range communication and negligible energy consumption. While LoRaWAN is considered the open standard with the highest level of acceptance from both industry and academia, it is its random access protocol (Aloha) that limits its capacity in large-scale deployments to some extent. Although time-slotted scheduling has proved to alleviate certain scalability limitations, the constrained nature of end nodes and their application-oriented requirements significantly increase the complexity of time-slotted network management tasks. To shed light on this matter, a multi-agent network management system for the on-demand allocation of resources in end-of-life monitoring applications for remanufacturing is introduced in this work. It leverages LoRa’s spreading factor orthogonality and network-wide knowledge to increase the number of nodes served in time-slotted monitoring setups. The proposed system is validated and evaluated for end-of-life monitoring where two representative end-node distributions were emulated, with the achieved network capacity improvements ranging from 75.27% to 249.46% with respect to LoRaWAN’s legacy operation. As a result, the suitability of different agent-based strategies has been evaluated and a number of lessons have been drawnaccording to different application and hardware constraints. While the presented findings can be used to further improve the explainability of the proposed models (in line with the concept of eXplainable AI), the overall framework represents a step forward in lightweight end-of-life condition monitoring for remanufacturing.

Reconfigured Self-Referenced MRAM Cell Supporting 2.527-fJ/bit Reading and In Situ Multiplication

The development of nonvolatile memories (NVM) based normally-off, instantly-on applications is hindered by high energy consumption for data readout. In particular, magnetic random access memory (MRAM) is a promising candidate for embedded non-volatile memory for its high cell density, high endurance and compatibility with the CMOS process. However, considering the low tunneling magnetoresistance ratio (TMR) of standard 1T1MTJ bit-cell, the state-of-art voltage sensing amplifiers (VSA) and current sensing amplifiers (CSA) need the direct current path to build considerable voltage or current difference. This paper demonstrates a reconfigured latch-based 4T2MTJ bit-cell achieving 2.527-fJ/bit ultra-low-power data readout at 0.4-V read voltage. Simulated results show that 2-ns latency can be realized with a traditional VSA and parasitic capacitance. The proposed 4T2MTJ bit-cell can perform in-situation (in-situ) multiplication between stored data and input with negligible extra energy consumption.

Bi-functional carbon foam/TiO2 composite absorber for solar steam generation and photocatalytic purification

Due to the population growth and environmental issues, the demand for clean water is still increasing. To address this problem, harvesting water from organic sewage by solar driven water evaporation is emerging as one of the most promising technologies with its attractive efficiency and negligible energy consumption. However, it still faces a typical defect as pollutants depositing on absorber, which usually limits the recyclability and causes a secondary pollution. In this work, we have provided a composite absorber by uniting carbon foam and nanostructured TiO2 for wastewater purification with dual function of solar steam generation and photocatalytic degradation. The composites selectively utilize the whole solar spectrum, in which low energy visible-near infrared (NIR) photons are transformed into heat by carbon foam, while high energy ultraviolet (UV) photon is converted into redox electron-hole pairs via TiO2. Besides, the combination of these two functional units reveals a synergistic incentive effect between photothermal and photocatalytic processes, based on the alleviated pollutant deposition and enhanced carrier separation. Under a simulated light of 1 kW m−2, the composite ultimately achieves a stable solar steam evaporation rate of 1.0334 kg m−2h−1, combining with a photocatalytic activity of degrading 87 % of MB after 2.5 h illumination.

Ising-Traffic: Using Ising Machine Learning to Predict Traffic Congestion under Uncertainty

This paper addresses the challenges in accurate and real-time traffic congestion prediction under uncertainty by proposing Ising-Traffic, a dual-model Ising-based traffic prediction framework that delivers higher accuracy and lower latency than SOTA solutions. While traditional solutions face the dilemma from the trade-off between algorithm complexity and computational efficiency, our Ising-based method breaks away from the trade-off leveraging the Ising model's strong expressivity and the Ising machine's strong computation power. In particular, Ising-Traffic formulates traffic prediction under uncertainty into two Ising models: Reconstruct-Ising and Predict-Ising. Reconstruct-Ising is mapped onto modern Ising machines and handles uncertainty in traffic accurately with negligible latency and energy consumption, while Predict-Ising is mapped onto traditional processors and predicts future congestion precisely with only at most 1.8% computational demands of existing solutions. Our evaluation shows Ising-Traffic delivers on average 98X speedups and 5% accuracy improvement over SOTA.

Electrospun chitosan nanofiber constructing superhigh-water-flux forward osmosis membrane

Forward osmosis (FO) technology exhibits great potential in seawater desalination and wastewater treatment due to its negligible energy consumption and high antifouling, however, the weak desalination capability, especially low water flux, remains challenging. Herein, a cost-effective and high-desalination-performance chitosan (CS)-based FO membrane is developed via coupling the electrospinning CS nanofibers and interfacial-polymerized polyamide (PA). The electrospun nanofibers construct the porous and hydrophilic CS layer with the large pore-diameter of ~274 nm and low thickness of ~10 μm, enabling the effective transport of water molecules, specifically, a superhigh water flux of 107.53 LMH at a low salt-water ratio of 0.24 g·L−1. In addition, such superior desalination performance of the as-prepared FO membrane is universal for the various salt species and concentrations. Our CS nanofiber-based membrane with the high separation capability of water-salt, desirable antibacterial activity, as well as the low cost, offers a roadmap toward the sustainable membrane materials.

First Principles Study of Double Boron Atoms Supported on Graphitic Carbon Nitride (g-C3N4) for Nitrogen Electroreduction

Electrocatalytic reduction of N2 provides a clean, sustainable way for NH3 production. Efficient catalysts thus play a key role but remain a long-term challenge. In this study, the catalytic activity of double boron supported on graphitic carbon nitride (g-C3N4) for a N2 reduction reaction (NRR) is explored by density functional theory (DFT) calculations. Our results show that double boron atoms embedded in g-C3N4 with coordination of four N atoms and two boron atoms exhibits an excellent NRR performance with negligible energy consumption for adding hydrogen to *N2, while a moderate ΔG of 0.58 eV for the formation of the second NH3 suggests this catalyst is a potential candidate for N2 fixation.

Latency-Aware Task Scheduling for IoT Applications Based on Artificial Intelligence with Partitioning in Small-Scale Fog Computing Environments.

The Internet of Things applications have become popular because of their lightweight nature and usefulness, which require low latency and response time. Hence, Internet of Things applications are deployed with the fog management layer (software) in closely located edge servers (hardware) as per the requirements. Due to their lightweight properties, Internet of Things applications do not consume many computing resources. Therefore, it is common that a small-scale data center can accommodate thousands of Internet of Things applications. However, in small-scale fog computing environments, task scheduling of applications is limited to offering low latency and response times. In this paper, we propose a latency-aware task scheduling method for Internet of Things applications based on artificial intelligence in small-scale fog computing environments. The core concept of the proposed task scheduling is to use artificial neural networks with partitioning capabilities. With the partitioning technique for artificial neural networks, multiple edge servers are able to learn and calculate hyperparameters in parallel, which reduces scheduling times and service level objectives. Performance evaluation with state-of-the-art studies shows the effectiveness and efficiency of the proposed task scheduling in small-scale fog computing environments while introducing negligible energy consumption.

Effects of pre-treatments on the filtration performance of ultra-low pressure gravity-driven membrane in treating the secondary effluent: Flux stabilization and removal improvement

Gravity-driven membrane (GDM) filtration is a promising technology for decentralized water and wastewater treatment due to its negligible energy consumption and low maintenance. This study investigated the flux development and permeate quality of the GDM process in treating the secondary effluent, as well as the influences of two practical pre-treatments, including GAC filter and microfiltration (MF), on its filtration performance. The results indicated that the GAC filter could significantly enhance the removal of organic contaminants (i.e., DOC, UV254, and fluorescent substances), with an average removal of 70 % for DOC and 80 % for UV254, respectively. Meanwhile, GAC/GDM systems exhibited a superior removal performance of manganese and sulfamethoxazole (SMX) for more than 95 % and 80 %, respectively. Furthermore, the optimized mixing ratio between GAC permeates and secondary effluent was 7:3 (GAC7:3) to reach a flux level close to the GDM control and achieve a superior permeate water quality than the other processes. The flux of GDM control could stabilize at the level of 5.47 ± 0.71 L m−2h−1 during long-term filtration. Unexpectedly, limited flux improvements were observed with the pre-treatments, and the stable flux of GAC/GDM and MF/GDM was close to or slightly lower than the GDM control. In GDM control, a rough, heterogeneous, and porous biocake layer was observed on the membrane surface. By contrast, the cake layer structures in the GDM system coupled with pre-treatments appeared denser, compact, and homogeneous, resulting in relatively lower stable flux relative to the GDM control. The economic assessment indicated that the optimized process of GAC7:3 showed a significant cost advantage in the small and medium-scale water reuse scenarios. These findings were beneficial to develop new energy-saving membrane-based technology for the reclamation of wastewater.

Steer by Image Technology for Intelligent Reflecting Surface Based on Reconfigurable Metasurface with Photodiodes as Tunable Elements

Lately, metasurface has become an essential and promising component in implementing Intelligent Reflecting Surface (IRS) for 5G and 6G. A novel method that simplifies the ability to reconfigure the metasurface is presented in this paper. The suggested technology uses a PIN photodiode as a tuning element. The desired image is projected on the metasurface’s backside, where the PIN photodiodes are placed and reconfigures the metasurface. The projected image’s color and intensity pattern influence the PIN photodiode’s junction capacitance, which leads to local reflection phase control. This enables the required pattern reflection phase distribution to manipulate the reflection beam, for example, 2D beam steering or focusing, and any other beam forming combination, instead of wiring many digital-to-analog converters (DACs) or FPGA outputs, which bias the standard tuning element such as PIN diode or varactor using a complex RF circuit. Using a PIN photodiode as a tunable element instead of a varactor diode, PIN diode, Liquid Crystal and MEMS allows the changing of the internal junction capacitance without direct contact and thus continuously controlling the reflection phase. In addition, an open circuit work mode with negligible energy consumption can be obtained. This technology can be used to implement metasurface based on discrete or continuous phases and is called Steer by Image (SBI). A full description of the SBI technology using PIN photodiode is presented in this paper.

Plant-based amyloids from food waste for removal of heavy metals from contaminated water

Water pollution is one of the major global threats brought about by industrial, agricultural, and any other anthropogenic activity. Heavy metals represent a large group of water pollutants that can accumulate in the human body, causing cancer and mutagenic diseases. Technologies currently used to treat polluted wastewaters of heavy metals employ chemical, ion-exchange, and membrane purification methods. However, these techniques are energy-intensive due to high pressure and power requirements for membrane-based technologies, or highly selective, as in ion-exchange resins, making drinking water less affordable in developing countries. In this study, plant amyloid-carbon membranes consisting of sunflower and peanut amyloid fibrils were fabricated through a green and sustainable process and were used to remove toxic heavy metal pollutants to drinkable standards with negligible energy consumption. Protein-rich sunflower and peanut meals serve as low-cost raw materials, from which proteins were extracted, isolated, and self-assembled into functional amyloid fibrils for heavy metal removal. These amyloid fibrils were incorporated into hybrid carbon/amyloid membranes and used to filer Pt-, Cr-, and Pb-containing water to produce water of drinkable standards containing < 10 ppb heavy metals. This process can easily be upscaled due to its simplicity and minimal use of chemical reagents, pointing towards the future of low-cost yet efficient water treatment technologies.

Mechanistic Investigation of Electrostatic Field-Enhanced Water Evaporation.

Investigations on external electrostatic field (EEF)‐enhanced liquid water evaporation have been reported decades ago, which suggest that molecular alignment and polarization tuned by EEF accelerating the phase change process could be responsible for EEF‐enhanced water evaporation. However, a detailed study revealing the role of EEF in altering the intermolecular and intramolecular water structure is lacking. Herein, an EEF is proved to tune water state by accelerating the thermal movement of water molecules, lowering the molecular escaping energy, and loosening the hydrogen bond structure. The detailed mechanisms and field interactions (heat and electrostatic) are investigated by in situ Raman characterizations and molecular dynamic simulations, which reveal that an EEF can effectively reduce the free energy barrier of water evaporation and then increase the evaporated water molecule flux. As a proof of concept, an EEF is integrated into an interfacial two‐dimentional solar steam generator, enhancing the efficiency by up to 15.6%. Similar to a catalyst lowing activation energy and enhancing kinetics of a chemical reaction, the EEF enhances water state tuning, lowers evaporation enthalpy, and then boosts steam generation rate with negligible additional energy consumption, which can serve as a generic method for water evaporation enhancement in water harvesting, purification, and beyond.

UAVs as an Intelligent Service: Boosting Edge Intelligence for Air-Ground Integrated Networks

The air-ground integrated network is a key component of future sixth generation (6G) networks to support seamless and near-instant super-connectivity. There is a pressing need to intelligently provision various services in 6G networks, which however is challenging. To meet this need, in this article, we propose a novel architecture called UaaS (UAVs as a Service) for the air-ground integrated network, featuring UAV as a key enabler to boost edge intelligence with the help of machine learning (ML) techniques. We envision that the proposed UaaS architecture could intelligently provision wireless communication service, edge computing service, and edge caching service by a network of UAVs, making full use of UAVs' flexible deployment and diverse ML techniques. We also conduct a case study where UAVs participate in the model training of distributed ML among multiple terrestrial users, whose result shows that the model training is efficient with a negligible energy consumption of UAVs, compared to the flight energy consumption. Finally, we discuss the challenges and open research issues in the UaaS.

A perceptron-based replication scheme for managing the shared last level cache

The shared last level cache (SLLC), which provides large effective cache capacity, is widely adopted in modern chip multiprocessors (CMPs). But, long on-chip access latency in the SLLC is a key problem that hurts system performance. Replication is an effective way to relieve this problem through storing a replica of L1 victims in the near local LLC slice. However, previous replication schemes either blindly create replicas based on no feature of cache blocks or select replicas based on a single feature (such as data type, access count, etc.), which will affect the replication accuracy and limit the system performance improvements. In this paper, according to the successful application of machine learning (ML) in the field of computer architecture optimization in recent years, we develop a novel perceptron-based replication scheme (PBR) for effectively managing the SLLC in CMPs. Unlike existing single-feature-based schemes, this scheme effectively combines four features related to the reuse behavior of L1 victims, which are address (Addr), program counter (PC), data type (DT), and access count (AC), through perceptron to facilitate the accuracy of replica selection. Experimental results show that compared with the two previously proposed single-feature-based replication schemes: ASR and LADR, PBR decreases the execution time by 6.59% and 18.27%, and reduces the network traffic by 10.35% and 13.18% respectively with negligible energy consumption, hardware and area overhead.

Microwave dehydration of potato slices and assessment of energy efficiency

The dehydration parameters (thickness, mass load, and power level) statistically significantly (p&lt;0.05) affect the microwave dehydration of potato slices. Potato slices with thicknesses of 3, 6, and 9 mm were dehydrated as monolayers at different mass loads (1.00, 0.63, and 0.38 kg m-2) and microwave power levels (80, 240 W). The optimal model of potato slices with a 3 mm thickness, 0.38 kg m-2 mass load, dehydrated on 240 W, had the shortest dehydration time (15 minutes), the most negligible energy consumption (0.064 kWh), and the most insignificant emission of carbon dioxide (0.063 kg). The model of potato slices of 9 mm slice thickness dehydrated on 240 W, with 0.38 kg m-2 mass load, showed the highest resistance to mass transfer (the maximum effective moisture diffusivity 1.1847 × 10-7 ± 2.6080 × 10-9 m2 s-1). The average activation energy for all models was determined to be 11.635 W g-1. The thinner potato slices showed better results in dehydration time and energy consumption and good moisture diffusivity.

Employ DBSCAN and Neighbor Voting to Screen Selective Forwarding Attack Under Variable Environment in Event-Driven Wireless Sensor Networks

In the event-driven wireless sensor networks (EWSNs), the event of interests occurs irregularly and at random in the network. Then, sensor nodes near the event sense the event and send out data packets of the event. Next, router nodes (RNs) forward those packets to the sink node (SN) by multi-hop communications. Compromised RNs would become malicious and launch selective forwarding attacks by dropping part of or all the packets from other nodes. On the other hand, a harsh environment makes the channel poor, so the routing nodes under a harsh environment have low packet forwarding rates because they sometimes have to give up forwarding the current packets after many tries to forward them due to poor channel. If the malicious nodes' forwarding rates become close to those of nodes under a harsh environment, the schemes based on packet forwarding rates for detecting selective forwarding attack may fail because they cannot differentiate the low data packet forwarding rates resulting from the malicious behaviors or harsh environment. To solve this problem, we provide a combined scheme for detecting selective forwarding attack in wireless sensor networks (WSNs) under harsh environments. This scheme employs a data clustering algorithm (DCA) to screen the malicious nodes out by clustering their cumulative forwarding rates (CFRs) and designs a voting decision method to protect the nodes under a harsh environment from being judged as malicious nodes. The simulation results show that our scheme has a low false detection rate (FDR) of 1% and a low missed detection rate (MDR) of 5% respectively with negligible energy consumption in WSNs under a local variable harsh environment.

On Max–Min Throughput in Backscatter-Assisted Wirelessly Powered IoT

Backscatter communication can potentially find wide Internet of Things (IoT) applications because of its negligible energy consumption. Traditional backscatter communication relies on either dedicated radio-frequency (RF) sources, such as RF identification readers and power beacons, or ambient RF sources, e.g., TV and cellular signals. In this article, we study a backscatter-assisted wirelessly powered IoT system where devices can backscatter when nearby devices are actively transmitting via RF. Our objective is to determine the transmission schedule for all devices that maximizes the minimum system throughput. We formulate such a scheduling problem as a linear program for both linear and random networks. A key step in the formulation is to identify the groups of devices that can simultaneously backscatter without causing interference. Simulation results show that the max-min system throughput in linear and random networks can be increased by 46 and 180 times, respectively, by using the proposed backscatter-assisted schedule as compared with the traditional time-division multiple access (TDMA).