Transfer Bandwidth Research Articles

Load Balanced PIM-Based Graph Processing

Graph processing is widely used for many modern applications, such as social networks, recommendation systems, and knowledge graphs. However, processing large-scale graphs on traditional Von Neumann architectures is challenging due to the irregular graph data and memory-bound graph algorithms. Processing-in-memory (PIM) architecture has emerged as a promising approach for accelerating graph processing by enabling computation to be performed directly on memory. Despite having many processing units and high local memory bandwidth, PIM often suffers from insufficient global communication bandwidth and high synchronization overhead due to load imbalance. This article proposes GraphB, a novel PIM-based graph processing system, to address all these issues. From the algorithm perspective, we propose a degree-aware graph partitioning algorithm that can generate balanced partitioning at a low cost. From the architecture perspective, we introduce tile buffers incorporated with an on-chip 2D-Mesh, which provides high bandwidth for inter-node data transfer. Dataflow in GraphB is designed to enable computation–communication overlap and dynamic load balancing. In a PyMTL3-based cycle-accurate simulator with five real-world graphs and three common algorithms, GraphB achieves an average 2.2× and maximum 2.8× speedup compared to the SOTA PIM-based graph processing system GraphQ.

Ultra-wide dual-band Rydberg atomic receiver based on space division multiplexing radio-frequency chip modules

Detecting microwave signals over a wide frequency range has numerous advantages as it enables simultaneous transmission of a large amount of information, and access to more spectrum resources. This capability is crucial for applications such as microwave communication, remote sensing, and radar. However, conventional microwave receiving systems are limited by amplifiers and band-pass filters that can only operate efficiently in a specific frequency range. Typically, these systems can only process signals within a three-fold frequency range, which limits the data transfer bandwidth of the microwave communication systems. Developing novel atom-integrated microwave sensors, for example, radio frequency (RF)-chip coupled Rydberg atomic receiver, provides opportunities for a large working bandwidth of microwave sensing at the atomic level. Here, an ultra-wide dual-band RF sensing scheme is demonstrated by space-division multiplexing two RF-chip-integrated atomic receiver modules. The system can simultaneously receive dual-band microwave signals that span a frequency range exceeding 6 octaves (300 MHz and 24 GHz). This work paves the way for multi-band microwave reception applications within an ultra-wide range by RF-chip-integrated Rydberg atomic sensor.

Design optimization for microstrip antennas based on polymethyl methacrylate (PMMA) substrate and carbon nanotube (CNT) conductive material in sub-6 Ghz band

Abstract Background The rapid expansion of modern smart applications, demanding faster data transfer and extensive bandwidth, has prompted the development of new-generation networks like 5G and 6G. These networks encompass additional frequency bands such as sub-6 GHz, millimeter waves, and terahertz bands to meet the growing bandwidth requirements. However, despite the substantial bandwidth available in these bands, several challenges must be addressed to overcome unfavorable propagation characteristics. Moreover, numerous applications necessitate wireless devices with antennas that exhibit high flexibility and exceptional radiation responses, particularly when subjected to bending effects. This requirement highlights the importance of polymers-based antennas that can adapt to changing conditions while maintaining optimal performance. The present comprehensive study delves into the performance evaluation of rectangular and circular microstrip antennas utilizing PMMA (polymethyl methacrylate) polymer substrate with varying thicknesses. Results Notably, CNTs (Carbon Nanotubes) are employed as an alternative to traditional copper for the conductive part and ground plane. Both PMMA-based antennas, integrated with CNTs, exhibit a compact footprint of 27.8 × 47.8 × 1.5 mm3 for the circular antenna and 22.8 × 39.5 × 1.5 mm3 for the rectangular antenna. Impressively, the realized gain of both antennas surpasses 5 dBi, demonstrating robust performance in both flat and bending scenarios across different substrate thicknesses. Conclusions The rectangular antenna achieves a bandwidth of approximately 200 MHz, while the circular microstrip antenna showcase annotable bandwidth of 500 MHz. These exceptional outcomes position the two microstrip antennas as highly suitable for a diverse range of emerging applications within the sub-6 GHz band (the frequency range below 6 GHz in the radio spectrum). Thus, the combination of PMMA substrate, CNTs and the compact form factor of the antennas presents a compelling solution for meeting the demands of modern applications requiring efficient wireless communication with enhanced performance and bandwidth.

A New Method for Intelligent Multimedia Compression Based on Discrete Hartley Matrix

Multimedia data (video, audio, images) require storage space and transmission bandwidth when sent through social media networking. Despite rapid advances in the capabilities of digital communication systems, the high data size and data transfer bandwidth continue to exceed the capabilities of available technology, especially among social media users. The recent growth of multimedia-based web applications such as WhatsApp, Telegram, and Messenger has created a need for more efficient ways to compress media data. This is because the transmission speed of networks for multimedia data is relatively slow. In addition, there is a specific size for sending files via email or social networks, because much high-definition multimedia information can reach the Giga Byte size. Moreover, most smart cameras have high imaging resolution, which increases the bit rate of multimedia files of video, audio, and image. Therefore, the goal of data compression is to represent media (video, audio, images, etc.) as accurately as possible with the minimum number of bits (bit rate). Traditional data compression methods are complex for users. They require a high processing power for media data. This shows that most of the existing algorithms have loss in data during the process of compressing and decompressing data, with a high bitrate for media data (video, audio, and image). Therefore, this work describes a new method for media compression systems by discrete Hartley matrix (128) to get a high speed and low bit rate for compressing multimedia data. Finally, the results show that the proposed algorithm has a high-performance speed with a low bit rate for compression data, without losing any part of data (video, sound, and image). Furthermore, the majority of users of social media are satisfied with the data compression interactive system, with high performance and effectiveness in compressing multimedia data. This, in turn, will make it easier for users to easily send their files of video, audio, and images via social media networks.

Transfer Bandwidth Optimization for Multichannel Time-Correlated Single-Photon-Counting Systems Using a Router-Based Architecture: New Advancements and Results

Time-correlated single-photon counting (TCSPC) is a powerful technique for time-resolved measurement of fast and weak light signals used in a variety of scientific fields, including biology, medicine, and quantum cryptography. Unfortunately, given its repetitive nature, TCSPC is recognized as a relatively slow technique. In the last ten years, attempts have been made to speed it up by developing multichannel integrated architectures. Yet, for the solutions proposed thus far, the measurement speed has not increased proportionally to the number of channels, reducing the benefits of a multichannel approach. Recent theoretical studies and prototypes have shown that it is possible to implement a new multichannel architecture, so-called router-based architecture, capable of optimizing the efficiency of data transfer from the integrated chip to the data processor, increasing the overall measurement speed. However, the first implementations failed to achieve the theoretical results due to implementation flaws. In this paper, we present a new logic for the router-based architecture that can operate at the same laser frequency and solve the issues of the previous implementation. Alongside the new logic, we present a new integrated low-jitter delay line combined with a new method for timing-signal distribution that allows the proper management of the pixel timing information. The new implementation is a step closer to realizing a router-based architecture that achieves the expected theoretical results. Simulations and bench tests support the results here reported.

Optimising remote health care delivery in Antarctica: a review of the current capabilities utilised in the British Antarctic Territory

ABSTRACT Injury in Antarctica can have a significant impact when considering transfer timelines of several weeks. Medical support to the British Antarctic Territory (BAT) is provided by deployed healthcare professionals and the utilisation of “reach-back” with telemedicine. This is paired with robust training and familiarisation with a system of modularised deployed equipment. This paper examines the current telemedicine strategy, infrastructure modularisation, and influence from military practice by the British Antarctic Survey Medical Unit (BASMU) for medical care at extreme reach. Current telemedicine practices and utilisation, as well as modular equipment capabilities across the BAT were reviewed to provide an outline of care delivery. Requests varied from expert advice to remote supervision of clinical procedures. Integration of commercially available solutions enabled real-time display of patient physiology. The deployment of modular resources has improved equipment availability and greater standardisation between sites. The sending of case notes and digital x-rays has been generally sufficient but, when greater supervision was required, limited data transfer bandwidth was a challenge. An ongoing review of deployed equipment capabilities may also enhance the ease with which remote support can be offered but an uplift in telemedicine capability will likely require infrastructure upgrades to maintain data transfer from 8000 miles away.

Cryogenic electron ptychographic single particle analysis with wide bandwidth information transfer

Advances in cryogenic transmission electron microscopy have revolutionised the determination of many macromolecular structures at atomic or near-atomic resolution. This method is based on conventional defocused phase contrast imaging. However, it has limitations of weaker contrast for small biological molecules embedded in vitreous ice, in comparison with cryo-ptychography, which shows increased contrast. Here we report a single-particle analysis based on the use of ptychographic reconstruction data, demonstrating that three dimensional reconstructions with a wide information transfer bandwidth can be recovered by Fourier domain synthesis. Our work suggests future applications in otherwise challenging single particle analyses, including small macromolecules and heterogeneous or flexible particles. In addition structure determination in situ within cells without the requirement for protein purification and expression may be possible.

Dielectric Property Effect with Dielectric Constant for Millimeter Wave Antenna Design

The millimeter Wave transmission provides large bandwidth for data transfer which can satisfy the 5th generation mobile network architecture three main features: High data rate, Low latency, Mass Connectivity. Due to the high propagation attenuation characteristic of millimeter Wave, the Directivity of Antenna Array greatly increase the transmitting distance. The antenna dimensions are designed relative to wave length for better radiation performance. In millimeter Wave bands, the wave length is under millimeter scale which the antenna can be integrated to Package called AiP (Antenna in Package) This article will discuss the influence of substrate material property for antenna behavior. The Dielectric Constant (Dk) of air is closed to 1, use substrate with low DK for Antenna in Package results better expression to increase Gain and Wider bandwidth. However, the wave length is also affected by Dielectric Constant, since the wavelength is inversely proportional to the square root of the Dk, the substrate with high Dk reduces the AiP size obviously. In this paper, we will provide the comparison of antenna performance for Low/High Dk by Patch Antenna which consider single antenna and antenna array. In the simulation, the result shows the difference between High and Low Dk. When Dielectric Constant increases to 6 from 3, the width of single antenna can be reduced 30% and in contrast the peak gain is 15% higher with low Dk. For handheld devices such as smartphone, the internal space is limited, the size of antenna becomes critical. With the help of pattern design to enhance antenna performance. The high Dk material can be the choice of antenna for handheld devices.

Chiplet Integration by Die-to-Die Pillar-Suspended Bridge

Chiplet integration technology, which has the potential to scalable expand capabilities by integrating many integrated circuit (IC) chips with different manufacturing processes, structures, and functions while maintaining the data transfer bandwidth between IC chips, is attracting attention as a promising method to solve the problems such as slowdown of Moore’s law and the von Neumann bottleneck. Die-to-die bridge structures that connect dies via locally placed bridge chips with high-density interconnects is expected for scalable chiplet integration. We proposed novel die-to-die bridge architecture called “Pillar-Suspended Bridge (PSB)” that connects between chiplet and bridge via “MicroPillars” in molded form using panel-level manufacturing process. We successfully demonstrated proof-of-concept (PoC) through prototyping chiplets/bridge integrated module with 40 µm pitch MicroPillars using 300 mm × 300 mm panel-level manufacturing process. We confirmed fine-pitch die-to-die connection with MicroPillars and 20 µm thick bridge die by cross-sectional analysis of the module. We also evaluated warpage behavior of the PSB module in the reflow temperature range using shadow moiré method and confirmed warpage of less than 10 µm. Finally, we discussed scalability of this technology including external connections for Fan-Out RDL and optics.

Porting the WAVEWATCH III (v6.07) wave action source terms to GPU

Abstract. Surface gravity waves play a critical role in several processes, including mixing, coastal inundation, and surface fluxes. Despite the growing literature on the importance of ocean surface waves, wind–wave processes have traditionally been excluded from Earth system models (ESMs) due to the high computational costs of running spectral wave models. The development of the Next Generation Ocean Model for the DOE’s (Department of Energy) E3SM (Energy Exascale Earth System Model) Project partly focuses on the inclusion of a wave model, WAVEWATCH III (WW3), into E3SM. WW3, which was originally developed for operational wave forecasting, needs to be computationally less expensive before it can be integrated into ESMs. To accomplish this, we take advantage of heterogeneous architectures at DOE leadership computing facilities and the increasing computing power of general-purpose graphics processing units (GPUs). This paper identifies the wave action source terms, W3SRCEMD, as the most computationally intensive module in WW3 and then accelerates them via GPU. Our experiments on two computing platforms, Kodiak (P100 GPU and Intel(R) Xeon(R) central processing unit, CPU, E5-2695 v4) and Summit (V100 GPU and IBM POWER9 CPU) show respective average speedups of 2× and 4× when mapping one Message Passing Interface (MPI) per GPU. An average speedup of 1.4× was achieved using all 42 CPU cores and 6 GPUs on a Summit node (with 7 MPI ranks per GPU). However, the GPU speedup over the 42 CPU cores remains relatively unchanged (∼ 1.3×) even when using 4 MPI ranks per GPU (24 ranks in total) and 3 MPI ranks per GPU (18 ranks in total). This corresponds to a 35 %–40 % decrease in both simulation time and usage of resources. Due to too many local scalars and arrays in the W3SRCEMD subroutine and the huge WW3 memory requirement, GPU performance is currently limited by the data transfer bandwidth between the CPU and the GPU. Ideally, OpenACC routine directives could be used to further improve performance. However, W3SRCEMD would require significant code refactoring to make this possible. We also discuss how the trade-off between the occupancy, register, and latency affects the GPU performance of WW3.

Energy Efficiency in Wireless Sensor Networks using Advanced LEACH Protocol

The wireless sensor network is the best research subject matter with huge applications in various domains. By and large, a wireless sensor network involves hundreds to thousands of sensor nodes, which transmit and communicate with each other by the utilization of radio transmissions or signals. A portion of the difficulties that exist in the sketch of wireless sensor networks are confined computation power, storage capacity, battery, and transmission transfer speed or bandwidth. To determine these issues, clustering and routing algorithm have been introduced. Clustering and routing processes are viewed as an optimization issue in wireless sensor networks which can be settled by the utilization of swarm intelligence-based approaches. This article presents an original multitude of insight-based grouping and multihop routing protocol for wireless sensor report giving a new swarm optimization technique applied for picking the cluster heads and organizing the cluster capably. Then, at that point, the grey wolf optimization algorithm-based routing process takes place to choose the ideal ways in the network. The introduced better particle swarm optimization-grey wolf optimization approach consolidates the advantages of both the clustering and routing processes which prompts the greatest energy efficiency and network lifetime. The proposed model is reproduced under a broad arrangement of experiments, and the outcomes are approved under a few measures. The acquired trial result exhibited the predominant qualities of the improved particle swarm optimization–grey wolf optimization method under all the test cases. It enhances LEACH protocols in terms of energy efficiency, network Lifetime, and throughput. Maintaining the Energy Efficiency of a wireless sensor network has been a great concern nowadays. The main aim is to overcome the drawback of Improvement of the energy efficiency of a wireless sensor network. It may be improved with the performance of an optimization algorithm using swarm Intelligence. The novelty of the project is Network lifetime and Energy Consumption. The future of this project is to share the complete information which is present in the cluster head without any interruption

A Wideband Circulator Leakage Canceler for Retro-Directive RF System

This letter presents a wideband circulator leakage canceler (CLC) IC using a negative capacitance (NCAP). It enhances a conventional circulator with a wider bandwidth and high isolation for a full-duplex retro-directive short-range wireless power transfer (WPT) application. The proposed CLC consists of a directional coupler (DC), a hybrid coupler, and a low-pass filter network (LPFN) integrated with NCAP, which shows a considerable capability of suppressing the leakage power signal from transmitter (TX) output to receiver (RX) input. The CLC was fully implemented in the Samsung LR6LP 65-nm RF CMOS process and occupied the size of 2.1 mm <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\times0.89$ </tex-math></inline-formula> mm. The measurement of IC with a commercial circulator shows an average cancelation of 19 dB over the range of 225 MHz from 2.31 to 2.535 GHz. The input P1dB in band was measured at 16 dBm while the noise figure (NF) of antenna-receiver (ANT-RX) path increased by 5.2 dB at 2.5 GHz.

Galvanic-coupled Trans-dural Data Transfer for High-bandwidth Intra-cortical Neural Sensing.

A digital-impulse galvanic coupling as a new high-speed trans-dural (from cortex to the skull) data transmission method has been presented in this paper. The proposed wireless telemetry replaces the tethered wires connected in between implants on the cortex and above the skull, allowing the brain implant to be "free-floating" for minimizing brain tissue damage. Such trans-dural wireless telemetry must have a wide channel bandwidth for high-speed data transfer and a small form factor for minimum invasiveness. To investigate the propagation property of the channel, a finite element model is developed and a channel characterization based on a liquid phantom and porcine tissue is performed. The results show that the trans-dural channel has a wide frequency response of up to 250 MHz. Propagation loss due to micro-motion and misalignments is also investigated in this work. The result indicates that the proposed transmission method is relatively insensitive to misalignment. It has approximately 1 dB extra loss when there is a horizontal misalignment of 1mm. A pulse-based transmitter ASIC and a miniature PCB module are designed and validated ex-vivo with a 10-mm thick porcine tissue. This work demonstrates a high-speed and miniature in-body galvanic-coupled pulse-based communication with a data rate up to 250 Mbps with an energy efficiency of 2 pJ/bit, and has a small module area of only 26 mm2.

A processing-in-pixel-in-memory paradigm for resource-constrained TinyML applications

The demand to process vast amounts of data generated from state-of-the-art high resolution cameras has motivated novel energy-efficient on-device AI solutions. Visual data in such cameras are usually captured in analog voltages by a sensor pixel array, and then converted to the digital domain for subsequent AI processing using analog-to-digital converters (ADC). Recent research has tried to take advantage of massively parallel low-power analog/digital computing in the form of near- and in-sensor processing, in which the AI computation is performed partly in the periphery of the pixel array and partly in a separate on-board CPU/accelerator. Unfortunately, high-resolution input images still need to be streamed between the camera and the AI processing unit, frame by frame, causing energy, bandwidth, and security bottlenecks. To mitigate this problem, we propose a novel Processing-in-Pixel-in-memory (P2M) paradigm, that customizes the pixel array by adding support for analog multi-channel, multi-bit convolution, batch normalization, and Rectified Linear Units (ReLU). Our solution includes a holistic algorithm-circuit co-design approach and the resulting P2M paradigm can be used as a drop-in replacement for embedding memory-intensive first few layers of convolutional neural network (CNN) models within foundry-manufacturable CMOS image sensor platforms. Our experimental results indicate that P2M reduces data transfer bandwidth from sensors and analog to digital conversions by {sim },21times, and the energy-delay product (EDP) incurred in processing a MobileNetV2 model on a TinyML use case for visual wake words dataset (VWW) by up to mathord {sim },11times compared to standard near-processing or in-sensor implementations, without any significant drop in test accuracy.

Capacitive Power Transfer System With Integrated Wide Bandwidth Communication

This letter proposed a novel method for compensation network design used in capacitive power transfer (CPT) systems to achieve simultaneous power and data delivery. The proposed method combines both ladder filter and resonant circuit designs, allowing the existing double-sided CL compensation topology to exhibit different characteristics under the power channel and the communication channel. The proposed system significantly reduces the design complexity and increases the bandwidth for data transfer while maintaining the circuit resonance at the wireless power transfer frequency. The system is compatible with both amplitude shift keying (ASK) and frequency shift keying (FSK) modulation methods, passing the high-frequency data carrier with an approximately unity gain. Detailed circuit parameter design and system performance analysis are presented, and a new metric power to signal is introduced to analyze the effect of the data communication on the power transfer. A prototype CPT system rated at 10 W is built, which demonstrates a 81.2% dc–ac efficiency and a data transfer rate of 10 kbps by using either ASK or FSK modulation techniques.

Dynamic Deduplication Algorithm for Cross-User Duplicate Data in Hybrid Cloud Storage

The escalating growth of distributed big data in hybrid cloud storage architecture introduces a new set of challenges. Constantly, content enrichment puts pressure on capacity. Nonetheless, the explosion of user data places a significant strain on broadband and storage capacity. Consequently, many cloud storage providers will implement deduplication to compress data, reduce transfer bandwidth, and reduce cloud storage space. In cloud storage systems, it is a data compression and storage optimization method. By locating and removing redundant data, it can save storage space and bandwidth. An MTHDedup deduplication strategy based on the Merkle hash tree is presented in a hybrid cloud environment to address the issue of convergent encryption algorithms being susceptible to brute-force attacks and ciphertext computation time overhead. Merkle hash trees are constructed using additional encryption algorithms to generate encryption keys during file- and block-level deduplication, ensuring that generated ciphertexts are unpredictable. The method is effective against both internal and external brute-force attacks, thereby increasing data security. Our method reduces the computational burden of ciphertext generation and the key storage space, and the performance advantage increases with the number of privilege sets.

Conduction band structure of high-mobility organic semiconductors and partially dressed polaron formation.

The energy band structure provides crucial information on charge transport behaviour in organic semiconductors, such as effective mass, transfer integrals and electron-phonon coupling. Despite the discovery of the valence (the highest occupied molecular orbital (HOMO)) band structure in the 1990s, the conduction band (the lowest unoccupied molecular orbital (LUMO)) has not been experimentally observed. Here we employ angle-resolved low-energy inverse photoelectron spectroscopy to reveal the LUMO band structure of pentacene, a prototypical high-mobility organic semiconductor. The derived transfer integrals and bandwidths from the LUMO are substantially smaller than those predicted by density functional theory calculations. To reproduce this bandwidth reduction, we propose an improved (partially dressed) polaron model that accounts for the electron-intramolecular vibrational interaction with frequency-dependent coupling constants based on Debye relaxation. This model quantitatively reproduces not only the transfer integrals, but also the temperature-dependent HOMO and LUMO bandwidths, and the hole and electron mobilities. The present results demonstrate that electron mobility in high-mobility organic semiconductors is indeed limited by polaron formation.

Requirements and Hardware Limitations of High-Frame-Rate 3-D Ultrasound Imaging Systems

The spread of high frame rate and 3-D imaging techniques has raised pressing requirements for ultrasound systems. In particular, the processing power and data transfer rate requirements may be so demanding to hinder the real-time (RT) implementation of such techniques. This paper first analyzes the general requirements involved in RT ultrasound systems. Then, it identifies the main bottlenecks in the receiving section of a specific RT scanner, the ULA-OP 256, which is one of the most powerful available open scanners and may therefore be assumed as a reference. This case study has evidenced that the “star” topology, used to digitally interconnect the system’s boards, may easily saturate the data transfer bandwidth, thus impacting the achievable frame/volume rates in RT. The architecture of the digital scanner was exploited to tackle the bottlenecks by enabling a new “ring“ communication topology. Experimental 2-D and 3-D high-frame-rate imaging tests were conducted to evaluate the frame rates achievable with both interconnection modalities. It is shown that the ring topology enables up to 4400 frames/s and 510 volumes/s, with mean increments of +230% (up to +620%) compared to the star topology.

Synthesis of Parametrically Coupled Networks

We show that a common language can be used to unify the description of parametrically-coupled circuits--parametric amplifiers, frequency converters, and parametric nonreciprocal devices--with that of band-pass filter and impedance matching networks. This enables one to readily adapt network synthesis methods from microwave engineering in the design of parametrically-coupled devices having prescribed transfer characteristics, e.g. gain, bandwidth, return loss, and isolation. We review basic practical aspects of coupled mode theory and filter synthesis, and then show how to apply both, on an equal footing, to the design of multi-pole, broadband parametric and non-reciprocal networks. We supplement the discussion with a range of examples and reference designs.

Multiple-access relay stations for long-haul fiber-optic radio frequency transfer.

We report on the realization of a long-haul radio frequency (RF) transfer scheme by using multiple-access relay stations (MARSs). The proposed scheme with independent link noise compensation for each fiber sub-link effectively solves the limitation of compensation bandwidth for long-haul transfer. The MARS can have the capability to share the same modulated optical signal for the front and rear fiber sub-links, simplifying the configuration at the repeater station and enabling the transfer system to have the multiple-access capability. At the same time, we for the first time theoretically model the effect of the MARS position on the fractional frequency instability of the fiber-optic RF transfer, demonstrating that the MARS position has little effect on system's performance when the ratio of the front and rear fiber sub-links is around 1:1. We experimentally demonstrate a 1 GHz signal transfer by using one MARS connecting 260 and 280 km fiber links with the fractional frequency instabilities of less than 5.9 × 10-14 at 1 s and 8.5 × 10-17 at 10,000 s at the remote site and of 5.6 × 10-14 and 6.6 × 10-17 at the integration times of 1 s and 10,000 s at the MARS. The proposed scalable technique can arbitrarily add the same MARSs in the fiber link, which has great potential in realizing ultra-long-haul RF transfer.