Multilevel-cell phase-change memory

In the modern digital era of big data applications, there is an ever-increasing demand for higher memory capacity that is both reliable and cost effective. In the domain of non-volatile memory systems, Flash-based storage devices have dominated the consumer space for the past 15 years and have also entered the enterprise storage system in the past 2-3 years. However, with Flash memory devices facing serious scalability limits, there is an imminent need to explore the viability of other non-volatile memory technologies that can replace or complement Flash-based storage in the near future. Significant research efforts have been invested by various universities and research organization across the globe into realizing the so-called next-generation memories (NGMs). Phase-change memory is one such technology, which is viewed as the most promising candidate among the emerging technologies. Phase-Change Memory (PCM) technology is not new as the principle of storing information in chalcogenide-based materials was first explored in the 1960s. However, with the predominance of charge-based memory technology (DRAM, Flash storage, etc) thriving thanks to the technological advancements ofmetal-oxide semiconductor field-effect transistor (MOSFET) based devices, it was not until the late 90s that renewed interest in the class of phase-change materials was ignited. This is the same genre of materials as have been widely used in laser driven optical storage (rewritable CDs and DVDs) during the past 20 years or so. Although PCM chips have already been mass-produced by companies like Micron and Samsung, their capacity was limited as they were based on single-level cell (SLC) storage. The primary research focus of this thesis is on increasing the memory capacity by storing more than one bit of information per device, known as multilevel-cell (MLC). Achieving MLC capability is quite challenging, and we disclose some of our significant achievements in the realization of MLC PCM in the past few years. In this thesis, a comprehensive thermoelectric model is proposed for investigating the thermoelectrics physics in PCM device operation. With the increasing role of thermoelectrics at the smaller technology nodes, the proposed model provides valuable insights for a complete understanding of device operation. The model is validated by comparing the simulation results with experimental measurements. The model can also be used for fine-tuning the material properties, device design and geometry to improve the efficacy of these devices. In the second part of the thesis, the dominant reliability concerns in PCM technology are addressed, particularly focusing on MLC operation. We present a readout circuit for PCM specifically designed for drift resilience in MLC operation. Drift resilience is achieved through the use of specific non-resistance-based cell-state metrics which, in contrast to the traditional cell-state metric, i.e., the low-field electrical resistance, have built-in drift robustness. By employing novel MLC-enabling techniques, we succeeded in demonstrating for the first time, reliable 3-bits/cell memory density with a data retention of 1 week at temperatures ranging from 30±C to 80±C on devices that had been pre-cycled one million times.

Structure-aware Multi-view 3D Reconstruction of Dislocations in TEM with Message Passing Neural Networks

Okan Altingövde, Pascal Fua, Gulnaz Ganeeva, Cécile Hébert, Anastasiia Mishchuk, Emadeddin Oveisi

Efficient analysis of the three-dimensional (3D) shape and distribution of curvilinear crystal defects, namely dislocations, is an open research topic in material science and computer vision. In order to determine the structural and opto-electrical characteristics of the material, accurate 3D reconstructions of dislocations are required. In its current state, an established way to obtain 3D information of dislocations is electron tomography which relies on acquiring dozens of images from tilted sample for wide range of angles. Although tomography yields sufficient results in many cases, it is manually demanding to acquire, register and process dozens of images. In previous works [1, 2], it is shown that a classical stereo reconstruction approach may be applied as an alternative in order to increase reconstruction throughput while providing good performance. A later work [3] employing modern data-driven computer vision models showed that this stereo approach can be largely automatized and be used in various imaging conditions. In this work, we further improve the performance and memory consumption of previous convolutional neural network (CNN) based stereo reconstruction approach and extend it to be applicable in cases where more than two images are available. Our proposed multi-view approach directly incorporates the structural prior of dislocations to the model"s architecture and estimates 3D line segments approximating 3D shape of dislocations. In its core, our method has two stages which are structural connectivity estimation and depth estimation. In first stage, we employ a CNN to estimate affinity of sampled image locations on dislocation cores. This affinity information is used in the later depth estimation stage to eliminate the depth discontinuities on dislocations (Figure1). Unlike previous deep data-driven method, our new approach uses structural knowledge to separate depth estimations of different dislocation segments and yields more accurate reconstructions. In Figure 2, an example affinity estimation is shown for 4 different image locations on the same dislocation. It may be seen that dislocation points that are connected by dislocation segment in 3D results in higher affinity. This information is crucial to connect 3D point estimations and obtain linear structures in 3D.

2021

Rate allocation and optimized decoding in inter-session network coding

Eirina Bourtsoulatze

Recent advances in data processing and communication systems have led to a continuous increase in the amount of data communicated over today’s networks. These large volumes of data pose new challenges on the current networking infrastructure that only offers a best effort mechanism for data delivery. The emergence of new distributed network architectures, such as peer-to-peer networks and wireless mesh networks, and the need for efficient data delivery mechanisms have motivated researchers to reconsider the way that information is communicated and processed in the networks. This has given rise to a new research field called network coding. The network coding paradigm departs from the traditional routing principle where information is simply relayed by the network nodes towards the destination, and introduces some intelligence in the network through coding at the intermediate nodes. This in-network data processing has been proved to substantially improve the performance of data delivery systems in terms of throughput and error resilience in networks with high path diversity. Motivated by the promising results in the network coding research, we focus in this thesis on the design of network coding algorithms for simultaneous transmission of multiple data sources in overlay networks. We investigate several problems that arise in the context of inter-session network coding, namely (i) decoding delay minimization in inter-session network coding, (ii) distributed rate allocation for inter-session network coding and (iii) correlation-aware decoding of incomplete network coded data. We start by proposing a novel framework for data delivery from multiple sources to multiple clients in an overlay wireline network, where intermediate nodes employ randomized inter-session network coding. We consider networks with high resource diversity, which creates network coding opportunities with possibly large gains in terms of throughput, delay and error robustness. However, the coding operations in the intermediate nodes must be carefully designed in order to enable efficient data delivery. We look at the problem from the decoding delay perspective and design solutions that lead to a small decoding delay at clients through proper coding and rate allocation. We cast the optimization problem as a rate allocation problem, which seeks for the coding operations that minimize the average decoding delay in the client population. We demonstrate the validity of our algorithm through simulations in representative network topologies. The results show that an effective combination of intra- and inter-session network coding based on randomized linear coding permits to reach small decoding delays and to better exploit the available network resources even in challenging network settings. Next, we design a distributed rate allocation algorithm where the users decide locally how many intra- and inter-session network coded packets should be requested from the parent nodes in order to get minimal decoding delay. The capability to take coding decisions locally with only a partial knowledge of the network statistics is of crucial importance for applications where users are organized in dynamic overlay networks. We propose a receiver-driven communication protocol that operates in two rounds. First, the users request and obtain information regarding the network conditions and packet availability in their local neighborhood. Then, every user independently optimizes the rate allocation among different possible intra- and inter-session packet combinations to be requested from its parents. We also introduce the novel concept of equivalent flows, which permits to efficiently estimate the expected number of packets that are necessary for decoding and hence to simplify the rate allocation process. Experimental results indicate that our algorithm is capable of eliminating the bottlenecks and reducing the decoding delay of users with limited resources. We further investigate the application of the proposed distributed rate allocation algorithm to the transmission of video sequences and validate the performance of our system using the NS-3 simulator. The simulation results show that the proposed rate allocation algorithm is successful in improving the quality of the delivered video compared to intra-session network coding based solutions. Finally, we investigate the problem of decoding the source information from an incomplete set of network coded data with the help of source priors in a finite algebraic field. The inability to form a complete decoding system can be often caused by transmission losses or timing constraints imposed by the application. In this case, exact reconstruction of the source data by conventional algorithms such as Gaussian elimination is not feasible; however, partial recovery of the source data may still be possible, which can be useful in applications where approximate reconstruction is informative. We use the statistical characteristics of the source data in order to perform approximate decoding. We first analyze the performance of a hypothetical maximum a posteriori decoder, which recovers the source data from an incomplete set of network coded data given the joint statistics of the sources. We derive an upper bound on the probability of erroneous source sequence decoding as a function of the system parameters. We then propose a constructive solution to the approximate decoding problem and design an iterative decoding algorithm based on message passing, which jointly considers the network coding and the correlation constraints. We illustrate the performance of our decoding algorithm through extensive simulations on synthetic and real data sets. The results demonstrate that, even by using a simple correlation model expressed as a correlation noise between pairs of sources, the original source data can be partially decoded in practice from an incomplete set of network coded symbols. Overall, this thesis addresses several important issues related to the design of efficient data delivery methods with inter-session network coding. Our novel framework for decoding delay minimization can impact the development of practical inter-session network coding algorithms that are appropriate for applications with low delay requirements. Our rate allocation algorithms are able to exploit the high resource diversity of modern networking systems and represent an effective alternative in the development of distributed communication systems. Finally, our algorithm for data recovery from incomplete network coded data using correlation priors can contribute significantly to the improvement of the delivered data quality and provide new insights towards the design of joint source and network coding algorithms.

EPFL2013

Towards real-world biomechanical analysis of performance and functional capacity using wearable sensors

Salil Apte

Today, wearable heart rate monitors are an integral part of runners' training sessions, with heart rate data routinely used to assess effort intensity and stress on the body. As athletes translate their physical capacity into performance on the field through their movement, biomechanical assessment can provide valuable information that complements physiological assessment. However, the potential of using biomechanical information in the evaluation of training sessions and standardized tests in practice remains largely untapped, partly because the assessment devices remain cumbersome to use and often require long post-processing, as well as programming skills. The proposed work aims to realize this potential by developing field methods for performance and capacity evaluation using portable inertial measurement units (IMUs) and Global Navigation Satellite System (GNSS) receivers.Running performance can be characterized by the ability to maintain appropriate running technique despite fatigue, while keeping the effort intensity prescribed by the coach, or planned as a pacing strategy. In this work, a systematic review was conducted to examine and synthesize the results of fatigue protocols in running, fol-lowed by continuous measurements during a competition to confirm the trends obtained from the review with data from the field and to measure the changes in running technique due to fatigue. In addition, models were developed to accurately estimate running power using foot-worn IMU over a range of speeds and inclines and validated using gold standard methods in the laboratory, to better characterize running intensity. The second part of this work consisted of investigating the ability of IMUs and GNSS to improve the evaluation of athletes in standardized tests, referred to as their functional capacity. Functional capacity is typically used by coaches to develop appropriate training loads for athletes. This work presented validated methods to instrument common functional tests with wearable sensors to measure the speed, agility, and endurance of athletes in the field. In addition, these methods enable the extraction and a deeper analysis of relevant biomechanical parameters that contribute to the measured capacity and help the sporting staff understand athletes' strengths and weaknesses in detail. All of the research conducted and methods developed in this work are based on various combinations of a minimal body-worn sensor setup with foot-worn IMUs and a single trunk-worn IMU-GNSS unit. The signal processing algorithms and models developed in this work allow the recorded signals to be translated into easily interpretable and actionable information. Based on this information, coaches and physical therapists can develop customized training programs that target the relevant parameters. The proposed sensor setups and methods have been used and validated in a variety of situations, such as pre-season testing of a professional soccer team, training sessions of elite sprinters, the Lausanne half-marathon race, etc., highlighting their potential for real-world application. I believe that this work will help pave the way towards a deeper understanding of the biomechanical contributions to performance in running and provide new tools for the development of personalized training and rehabilitation programs, with the goal of optimizing positive adaptation to training stimuli, thereby improving performance and reducing the incidence of injury.

EPFL2022