High-level Modulation Research Articles

Adopting automated bug assignment in practice — a longitudinal case study at Ericsson

[Context] The continuous inflow of bug reports is a considerable challenge in large development projects. Inspired by contemporary work on mining software repositories, we designed a prototype bug assignment solution based on machine learning in 2011-2016. The prototype evolved into an internal Ericsson product, TRR, in 2017-2018. TRR’s first bug assignment without human intervention happened in April 2019. [Objective] Our study evaluates the adoption of TRR within its industrial context at Ericsson, i.e., we provide lessons learned related to the productization of a research prototype within a company. Moreover, we investigate 1) how TRR performs in the field, 2) what value TRR provides to Ericsson, and 3) how TRR has influenced the ways of working. [Method] We conduct a preregistered industrial case study combining interviews with TRR stakeholders, minutes from sprint planning meetings, and bug-tracking data. The data analysis includes thematic analysis, descriptive statistics, and Bayesian causal analysis. [Results] TRR is now an incorporated part of the bug assignment process. Considering the abstraction levels of the telecommunications stack, high-level modules are more positive while low-level modules experienced some drawbacks. Most importantly, some bug reports directly reach low-level modules without first having passed through fundamental root-cause analysis steps at higher levels. On average, TRR automatically assigns 30% of the incoming bug reports with an accuracy of 75%. Auto-routed TRs are resolved around 21% faster within Ericsson, and TRR has saved highly seasoned engineers many hours of work. Indirect effects of adopting TRR include process improvements, process awareness, increased communication, and higher job satisfaction. [Conclusions] TRR has saved time at Ericsson, but the adoption of automated bug assignment was more intricate compared to similar endeavors reported from other companies. We primarily attribute the difference to the very large size of the organization and the complex products. Key facilitators in the successful adoption include a gradual introduction, product champions, and careful stakeholder analysis.

Learning to Cut via Hierarchical Sequence/Set Model for Efficient Mixed-Integer Programming.

Cutting planes (cuts) play an important role in solving mixed-integer linear programs (MILPs), which formulate many important real-world applications. Cut selection heavily depends on (P1) which cuts to prefer and (P2) how many cuts to select. Although modern MILP solvers tackle (P1)-(P2) by human-designed heuristics, machine learning carries the potential to learn more effective heuristics. However, many existing learning-based methods learn which cuts to prefer, neglecting the importance of learning how many cuts to select. Moreover, we observe that (P3) what order of selected cuts to prefer significantly impacts the efficiency of MILP solvers as well. To address these challenges, we propose a novel hierarchical sequence/set model (HEM) to learn cut selection policies. Specifically, HEM is a bi-level model: (1) a higher-level module that learns how many cuts to select, (2) and a lower-level module-that formulates the cut selection as a sequence/set to sequence learning problem-to learn policies selecting an ordered subset with the cardinality determined by the higher-level module. To the best of our knowledge, HEM is the first data-driven methodology that well tackles (P1)-(P3) simultaneously. Experiments demonstrate that HEM significantly improves the efficiency of solving MILPs on eleven challenging MILP benchmarks, including two Huawei's real problems.

A modular library for fast prototyping of solution-state nuclear magnetic resonance experiments

Abstract. We present a framework library (Modular Elements, ME) for the development of pulse sequences for Bruker spectrometers. It implements a two-level abstraction approach – the lower level comprises basic functional elements of pulse sequences and the higher one often-reused blocks comprising several evolution periods. The low-level abstractions reduce code duplication between variants of experiments such as hard-pulse and selective variants of individual NMR experiments. The high-level modules enable further reuse of pulse program code and aid in the construction of complex experiments. We show the library's functionality by presenting pulse programs that can be switched between standard and TROSY variants as well as hard and shaped pulses and that can seamlessly incorporate real-time homodecoupling. Adaptability is further demonstrated in a configurable 4D NOESY program.

Achievable Spectral Efficiency of Line-of-Sight MIMO Fixed Links with Compact Planar Antenna Arrays and Power Amplifier Limitations

Fixed Point-to-Point microwave wireless systems with high spectral efficiency are needed to meet the pervasive and increasing demand for capacity in back-haul networks of mobile radio systems. In this context, spatially multiplexed LoS-MIMO (Line-of-Sight Multiple Input Multiple Output) systems have been studied for about twenty years, particularly in the millimeter wave frequency bands (above 15 GHz). However, their deployment in real networks has been really limited, to the authors’ knowledge. This has been due to several factors, i.e. the practical possibility of using extremely high-level modulation formats (nowadays up to 8192-QAM), the joint use of co-channel dual polarization, and the availability of wider channel bands in the new high frequency ranges (e.g. E-Band). In addition, a crucial reason has been the difficulty of installing multiple antennas spaced apart in order to maximize the MIMO spatial multiplexing so providing the maximum capacity gain. This optimal antenna separation, even for the classical MIMO M×N\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$M \ imes N$$\\end{document} with M = 2 antennas at the receiver and N = 2 antennas at the transmitter, can be several meters, e.g. 5.71 m at 23 GHz on a 5 km link. In this article, we analyze the performance of LoS-MIMO systems where antenna separation is highly sub-optimal, for limiting the array size, and a satisfactory performance is made possible by the exploitation of specific bit loading and power allocation strategies and the setting of the working region of the RF transmitter power amplifiers to operate at a given Signal-to-Inter Modulation Distortion Ratio (SIMDR). The result is an overview of the advantages and drawbacks of compact LoS-MIMO from a wider perspective than in the existing literature, including fundamental aspects for the practical implementation of these systems. Performance is discussed in many cases of interest and compared with the state of the art SISO (Single Input Single Output) system.

Hardware and Firmware Design and Implementation of Twin 8-Bit and 32-Bit Microcontroller Boards for Research and Educational Applications

This document describes the design and implementation sequence of hardware and firmware components of two microcontroller boards of 8-bit and 32-bit architectures, which use similar form factors and share common input/output devices and recently developed web-based user interfaces. These developments are the result of over two decades of experimentation in educational and research measurement systems in renewable energy projects. The boards are intended to fill a gap between low-cost hobby systems and high-end research loggers, especially where international standards require control of uncertainties, as is the case with small wind turbine power curve assessments. The focus was to increase reliability in the hardware aspects and make software development easier by including user-friendly firmware and a set of in-house C language coding rules for new developments and refactoring of legacy code. Testing of the low-level driver sections and the use of automated tools such as Ceedling for higher-level module testing are also described.

Hierarchical Reinforcement Learning in Multi-Domain Elastic Optical Networks to Realize Joint RMSA

To improve the network scalability, a large elastic optical network is typically segmented into multiple autonomous domains, where each domain possesses high autonomy and privacy. This architecture is referred to as the multi-domain elastic optical network (MDEON). In the MDEON, the routing, modulation, and spectrum allocation (RMSA) for the inter-domain service requests are challenging. As a result, deep reinforcement learning (DRL) has been introduced recently where the RMSA policies are learned during the interaction of the DRL agents with the MDEON environment. Due to the autonomy of each domain in the MDEON, joint RMSA is essential to improve the overall performance. To realize the joint RMSA, we propose a hierarchical reinforcement learning (HRL) framework which consists of a high-level DRL module and multiple low-level DRL modules (one for each domain), with the collaboration of DRL modules. More specifically, for the inter-domain service requests, the high-level module obtains some abstracted information from the low-level DRL modules and generates the inter-domain RMSA decision for the low-level modules. Then the low-level DRL module gives the intra-domain RMSA decision and feeds back to the high-level module. The proposed HRL framework preserves the autonomy of each single domain while delivering effective overall network performance through the cooperation of the high-level and low-level DRL modules. Simulation results demonstrate that our proposed method outperforms previous approaches.

Semi-supervised semantic segmentation with cross teacher training

Convolutional neural networks can achieve remarkable performance in semantic segmentation tasks. However, such neural network approaches heavily rely on costly pixel-level annotation. Semi-supervised learning is a promising resolution to tackle this issue, but its performance still far falls behind the fully supervised counterpart. This work proposes a cross-teacher training framework with three modules that significantly improves traditional semi-supervised learning approaches. The core is a cross-teacher module, which could simultaneously reduce the coupling among peer networks and the error accumulation between teacher and student networks. In addition, we propose two complementary contrastive learning modules. The high-level module can transfer high-quality knowledge from labeled data to unlabeled ones and promote separation between classes in feature space. The low-level module can encourage low-quality features learning from the high-quality features among peer networks. In experiments, the cross-teacher module significantly improves the performance of traditional student–teacher approaches, and our framework outperforms state-of-the-art methods on benchmark datasets.

Precision tracking control of a dual-stage measuring machine

Abstract Modern production requires shorter measuring cycles of measuring machines, which can be achieved with highly dynamic references causing dynamic deviations of the actual tool-center-point (TCP) position. To minimize the TCP tracking error, the considered measuring machine is extended with a redundant axis and a modular control concept is proposed. For this dual-stage actuation setting, a higher-level reference allocation module exploits the resulting redundancy and yields suitable position references for the lower-level controlled subsystems. On the higher-level, two dual-stage control concepts are presented, yielding both significantly reduced tracking errors in experiments compared to using only the main axis. Furthermore, to deal with strongly spatially varying friction of the main axis of the considered measuring machine, its lower-level control system is improved.

LDM-Ex-FDM: A Novel Multi-Service Transmission Scheme for the ATSC 3.0 System

In order to improve system coverage performance, in this paper, a multi-service transmission scheme on the basis of the puncturing technique, namely layered division multiplexing extension frequency-division multiplexing (LDM-Ex-FDM), is proposed. The key idea of the proposed scheme is that the symbols punctured from the enhanced layer (EL) of fixed services will be independently transmitted on a certain number of subcarriers orthogonal to the LDM signal of the core layer (CL) and most of the EL. By doing so, the punctured symbols will be demodulated with a higher signal-to-noise ratio (SNR) on the receiver side and can be recovered well, thus improving the reception performance of the EL fixed services. Moreover, two puncturing strategies based on bit-interleaved coded modulation (BICM) symbols of the LDM-Ex-FDM scheme, namely equal interval symbol puncturing (EISP) and non-equal interval symbol puncturing (NEISP), are developed to improve the performance of the EL services. Simulation results are given to show that, when the puncturing rate in the LDM-Ex-FDM scheme is configured as 1/12, the performance of the fixed service can be improved by 1.8 dB, meanwhile the bit error rate (BER) performance of the mobile service is not affected. Thus, the overall system coverage performance can be improved. Furthermore, to reduce the computational complexity and the demodulation delay of fixed service carried in the EL, at the receiver, a direct interference cancellation (DIC) detector is proposed as well. Simulation results are given to show that, under high injection level and high modulation order, the proposed DIC scheme is able to achieve almost the same BER performance as that of the traditional successive interference cancellation (SIC) scheme while with lower computational complexity.

Mitigation of Nonlinear Effects using Machine Learning in Coherent Optical Access Networks

Introduction/Context: The use of coherent detection jointly with high-level modulation formats such as 16 and 64-QAM seems to be a convenient strategy to increment capacity of future optical access networks. However, coherent detection requires high complexity digital signal processing to mitigate different impairments. Objective: Mitigate signal distortions using nonsymmetrical demodulation techniques based on Machine Learning (ML) algorithms. Methodology: First, a single channel Nyquist m-QAM system at 28 and 32 Gbps was simulated in VPIDesignSuite software. Then, different signals modulated at 16 and 64-QAM were generated with different laser linewidth, transmission distances and launch powers. Two ML algorithms were implemented to carry out the demodulation of the generated signals. The performance of the algorithms was evaluated using the bit error rate (BER) in terms of different system parameters as laser linewidth, transmission distance, launch power and modulation format. Results: The use of ML allowed gains up to 2 dB in terms of optical signal-to-noise ratio at a BER value of for 16-QAM and 1.5 dB for 64-QAM. Also, the use of ML showed that it is possible to use a lower cost laser (100 kHz linewidth vs 25 kHz) with a better BER performance than using conventional demodulation. Conclusions: We showed that the use of both algorithms could mitigate nonlinear effects and could reduce computational complexity for future optical access networks.

Aquatic Animal Image Classification Technology Based on Transfer Learning and Data Augmentation

Yuan, H.; Zhang, S.; Qin, E., and Zhou, H., 2020. Aquatic animal image classification technology based on transfer learning and data augmentation. In: Hu, C. and Cai, M. (eds.), Geo-informatics and Oceanography. Journal of Coastal Research, Special Issue No. 105, pp. 129–133. Coconut Creek (Florida), ISSN 0749-0208.A method based on data augmentation and transfer learning methods is proposed to address the problems of complicated steps, low accuracy, and poor generalization of traditional aquatic animal image classification methods. First, a deep convolutional generative adversarial network is used to enhance the data of the sample. Second, based on modifying the source model's fully connected classification layer, the weights of high-level convolution modules are set to be trained for adaptive adjustment. Finally, using the accuracy on the validation set and training time as the evaluation indexes, the performance experiments were conducted on various network structures and a different proportion of trainable parameters. The experimental results show that when the portion of trainable parameters is around 75%, the accuracy of the image recognition model obtained by retraining can reach 97.9%, which can be improved by up to 36 percentage points compared with the support vector machine algorithm.

High-efficiency fabrication of computer-generated holograms in silica glass using a femtosecond Bessel beam

We demonstrated a method for computer-generated holograms fabricated by using a femtosecond Bessel beam. A high-efficiency point-by-point process was used to fabricate a longitudinal modified array with a period of 10 μm and a modified area with diameter of approximately 1.5 μm on transparent glass. Tuning the modified length of each area by changing the relative positions of the Bessel beam and sample allowed the phase profile of the incident light to be flexibly adjusted to generate an arbitrary far-field pattern. The approach was verified by inscribing a designed phase hologram in borosilicate glass. Twin images were eliminated by high-level modulation.

A Spectral and Energy Efficiency of Adaptive Modulation (AM) with Power Control in Wireless Sensor Network

This paper presents a combination of power control and adaptive modulation performance in wireless sensor networks (WSN). WSN comprised numerous smart and small sensor nodes with limited power, memory, and processing. Furthermore, there is an impairment wireless channel between the sensor node and the coordinator node. Energy and spectral efficiency are essential issues in WSN. These two targets can be reached by applying a combination of power control and adaptive modulation. This work proposes a power control system to determine the modulation level through feedback SNR value of instantaneous wireless channel condition. The modulation level continuously changed based on the wireless channel condition. This adaptive modulation increases spectral efficiency. A better channel condition will drive to switch into a higher modulation level and vice versa. The result shows that the existence of power control can increase by about 9% of spectral efficiency than a system without power control. The system without power control just has a 91 % of selection probability, while the system equipped with power control creates a 100 % selection probability at the same of modulation level (16 QAM). Besides, combining power control with adaptive modulation also adjusts transmitting power at sensor nodes so that the effective use of transmitting power at the sensor node can be achieved. Combining power control and adaptive modulation can decrease the power consumption, enlarge the spectral efficiency with specific target bit error rate (BER), reduce the interference level, and significantly improve the WSN system performance.

The Effect of Feedback Delay on the Performance of Closed-Loop Rotate Modulations on OFDM/OFDMA Systems

In this paper, in order to improve the performance of the OFDM/OFDMA system, novel closed-loop rotate modulation is proposed by utilizing feedback channel state information in the complex values. The basic idea of the proposed rotate modulation is the channel equalizer process, which is applied at the transmitter side; therefore, it does not need to insert guard time to anticipate the problem of channel delay spread. With the purpose of distortionless-transmission, in this paper, the rotate modulation equation has been derivated. Based on the simulation results on this proposed system, when the feedback delay has been zero, the resulting system performance has been similar to the one on the Additive White Gaussian Noise channel. In this ideal condition, the received signal has been similar to the transmitted signal although in real condition the process of transmitting the feedback signal will always result in a delay. Simulation results at low-level modulation such as Binary Phase Shift Keying and Quarternary Phase Shift Keying; the system bit error ratio performance has not shown degradation with the feedback delay. However, for higher level modulation such as 16-Quadrature Amplitude Modulation and 64-Quadrature Amplitude Modulation, the feedback delay has caused degradation on the BER curve.

A Survey of Computer Vision Methods for 2D Object Detection from Unmanned Aerial Vehicles.

The spread of Unmanned Aerial Vehicles (UAVs) in the last decade revolutionized many applications fields. Most investigated research topics focus on increasing autonomy during operational campaigns, environmental monitoring, surveillance, maps, and labeling. To achieve such complex goals, a high-level module is exploited to build semantic knowledge leveraging the outputs of the low-level module that takes data acquired from multiple sensors and extracts information concerning what is sensed. All in all, the detection of the objects is undoubtedly the most important low-level task, and the most employed sensors to accomplish it are by far RGB cameras due to costs, dimensions, and the wide literature on RGB-based object detection. This survey presents recent advancements in 2D object detection for the case of UAVs, focusing on the differences, strategies, and trade-offs between the generic problem of object detection, and the adaptation of such solutions for operations of the UAV. Moreover, a new taxonomy that considers different heights intervals and driven by the methodological approaches introduced by the works in the state of the art instead of hardware, physical and/or technological constraints is proposed.

Image Segmentation Algorithms for Banana Leaf Disease Diagnosis

Identification and classification of leaf diseases in banana crop are an important requirement for farmers to diagnose and to get proper remedies for the pest and disease infection. Development of an automated system using image processing for leaf disease identification reduces time, cost and mainly supports to increase the productivity of banana fruit. In this process of automation, image segmentation is a key component that is required to analyze the image and to extract information from it. Image segmentation is a low-level module of image processing used to segregate the required object from an image for further analysis. The performance accuracy of image segmentation module determines the success of higher-level module of image processing. Therefore, to select an appropriate segmentation method for leaf analysis, different segmentation methods like adaptive thresholding, canny, color segmentation, fuzzy C-means, geodesic, global thresholding, K-means, log, multithresholding, Prewitt, region growing, Robert, Sobel and zero crossing are analyzed and compared in this paper. The quantitative matrices such as mean square error (MSE), peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) are considered to measure the performance of different segmentation methods. The results showed that geodesic method had significantly lower MSE value (6610), PSNR value (6608) and higher SSIM value (0.196) than all other methods. It is concluded that geodesic method is better for segmentation of banana leaf disease images.

LoRaFarM: A LoRaWAN-Based Smart Farming Modular IoT Architecture.

Presently, the adoption of Internet of Things (IoT)-related technologies in the Smart Farming domain is rapidly emerging. The ultimate goal is to collect, monitor, and effectively employ relevant data for agricultural processes, with the purpose of achieving an optimized and more environmentally sustainable agriculture. In this paper, a low-cost, modular, and Long-Range Wide-Area Network (LoRaWAN)-based IoT platform, denoted as “LoRaWAN-based Smart Farming Modular IoT Architecture” (LoRaFarM), and aimed at improving the management of generic farms in a highly customizable way, is presented. The platform, built around a core middleware, is easily extensible with ad-hoc low-level modules (feeding the middleware with data coming from the sensors deployed in the farm) or high-level modules (providing advanced functionalities to the farmer). The proposed platform has been evaluated in a real farm in Italy, collecting environmental data (air/soil temperature and humidity) related to the growth of farm products (namely grapes and greenhouse vegetables) over a period of three months. A web-based visualization tool for the collected data is also presented, to validate the LoRaFarM architecture.

On The Quality Measurements of RF Signals Using GNU Radio for Software Defined Radio Implementations in Future RF Networks

This paper presents the evaluation of the signal quality based on the Error Vector Magnitude (EVM) online measurements for different radio access modulations implemented over the GNU radio platform for Software Defined Radio (SDR) developments. It is demonstrated that when a transmitted symbol is unknown by the receiver, the EVM can be calculated by measuring the closest distance between the received symbol and the ideal one of the analyzed digital constellation. The implementation of the EVM measurements through software development for multilevel digital modulations has enabled obtaining and assessing the Signal-to-Noise Ratio (SNR) and Bit Error Rate (BER) values. Results show that the SNR measurements accordingly depicts the expected behavior of the analyzed digital modulations and BER floor appears at high level modulation formats due to the way the EVM itself has been obtained.

Impact of Frequency Shift on Nonlinear Compensation Using Optical Phase Conjugation for M-QAM Signals

Nonlinear compensation using optical phase conjugation (OPC) have been considered a promising technique to increase the reach of high-speed fiber-optic transmission systems. OPC-based nonlinear compensation employs an optical phase conjugation located at a middle of the fiber link to generate a complexed conjugated signal with respect the signal in the first half of the link for propagation in the second half. OPC technique assumes a symmetry for signal propagating in the first and second half to obtain a perfect nonlinear and chromatic dispersion. However, as most of practical OPC schemes are realized by nonlinear effects such as four-wave mixing or a combination of second-harmonic generation and difference frequency generation, the frequency shift induced by OPC affects the signal symmetrical requirement for nonlinear compensation because the chromatic dispersion is different for the first and second half transmissions. In this paper, we investigate the impact of frequency shift on the nonlinear compensation using OPC for high symbol rate, high level modulation format signals. This will be important to understand the tolerance of the OPC techniques against such a practical condition for actual system implementations.

Integrated Platform for the Analysis and Design of Tall Buildings for Wind Loads

Abstract The Romanian, as well as other wind design codes for building structures, provides with limited degree of accuracy, the aerodynamic loads distribution on buildings up to 200 meters tall in an equivalent static approach (ESWL). For tall wind-sensitive building structures, especially for those with irregular shapes, most of the codes or standards recommend for design to use pressure data recorded in the wind tunnel. The ESWL approach is however used as reference estimation and structural first phase design. Advances on experimental and computational capabilities, led in the past decade to a significant development of time-domain analysis framework, both for seismic and wind loads. While the major outcome for earthquake engineering practitioners is to select appropriate design input ground motions at a particular site, the wind engineering practitioners are facing numerical difficulties to handle large wind loading durations, especially dealing with nonlinear-induced effects. The paper presents a real-time integrated framework for the analysis and design of tall buildings to wind loads, based on the time-domain analysis tool, as a prerequisite for higher level modules as vulnerability, risk and loss estimation, and optimization analyses.