Analog Network Research Articles

Determination of the Effects of Transcutaneous Auricular Vagus Nerve Stimulation on the Heart Rate Variability Using a Machine Learning Pipeline.

Background:We are all aware of day-to-day healthy stress, but, when sustained for long periods, stress is believed to lead to serious physical and mental health issues.Materials and Methods:In this study, we investigated the potential effects of transcutaneous auricular vagus nerve stimulation (taVNS) on stress processing as reflected in the electrocardiogram (ECG)-derived biomarkers of stress adaptability. Stress reflecting biomarkers included a range of heart rate variability metrics: standard deviation of N-N intervals (SDNN), root mean squared of successive differences in heartbeat intervals (RMSSD), low-frequency component, high-frequency component and their ratio (LF, HF, and LF/HF).In addition, we created a machine learning model capable of distinguishing between the stimulated and nonstimulated conditions from the ECG-derive data from various subjects and states. The model consisted of a deep convolutional neural network, which was trained on R-R interval (RRI) data extracted from ECG and time traces of LF, HF, LF/HF, SDNN, and RMSSD.Results:Only LF/HF ratio demonstrated a statistically significant change in response to stimulation. Although the LF/HF ratio is expected to increase during exposure to stress, we have observed that stimulation during exposure to stress counteracts this increase or even reduces the LF/HF ratio. This could be an indication that the vagus nerve stimulation decreases the sympathetic activation during stress inducement.Our Machine Learning model achieved an accuracy of 70% with no significant variations across the three states (baseline, stress, and recovery). However, training an analogous neural network to identify the states (baseline, stress, and recovery) proved to be unsuccessful.Conclusion:Overall, in this study, we showed further evidence of the beneficial effect of taVNS on stress processing. Importantly we have also demonstrated the promising potential of ECG metrics as a biomarker for the development of closed-loop stimulation systems.

Packet-level and IEEE 802.11 MAC frame-level analysis for IoT device identification

In cyberspace, a large number of Internet of Things (IoT) devices from different manufacturers with hetero-geneous functionalities are connected together. It is challenging to identify all these devices in an IoT ecosystem. The situation becomes even more complicated when the devices come from the same manufacturer and of similar types due to their analogous network communication behaviour. In this paper, a device fingerprinting (DFP) approach based on a set of combined features from packet-level and frame-level has been proposed. A large number of features has been studied, and consequently, a suitable subset of features has been selected according to gain-ratio and device-specific features for DFP. Furthermore, experiments with different types of IoT devices in a laboratory environment to collect network traffic traces have been conducted and used to evaluate the performance of the proposed approach using the J48 (C4.5) algorithm. It has been shown that the proposed model is able to identify individual device types with 99.0% precision and 98.9% recall with the approach capable of classifying IoT devices coming from the same manufacturer and of similar types, with higher accuracy. These results are significant as it can be used as a security reinforcement tool towards increasing the security and resilience of IoT networks.

First Order Mixed Mode MOS-C All-Pass Frequency Selective Analog Network with Electronic Tuning

This work is intended to present a novel MOS-C design of first order frequency selective analog structure that plays essential role in phase equalization. The proposed idea employs 2 electronically tunable operational trans-conductance amplifiers, 7 MOS transistors forming active resistors and 1 grounded capacitor. Substantial flexibility to work in all 4 possible mode of operation enriches the uniqueness of proposed frequency selective structure. Non-ideal scenarios along with parasitic effects are also incorporated to explore real time performance of proposed structure. The emphasis on design has been enhanced by studying the effects of capacitor variations through Monte-Carlo analysis and the effects due to the temperature variations. Typical 0.18 µm CMOS process parameters are utilized in the verification of presented theoretical aspects through PSPICE simulation. To make room for the practicability of the proposed circuit, the experimental realization using commercially available ICs is also explored and included.
 HIGHLIGHTS
 
 A novel MOS-C design of first order frequency selective analog network is presented in this paper
 The proposed idea employs two electronically tunable operational trans-conductance amplifiers, seven MOS transistors forming active resistors and one grounded capacitor
 Non-ideal scenarios along with parasitic effects are also incorporated to explore real time performance of proposed structure
 18 µm CMOS process parameters are utilized in the verification of presented theoretical aspects through PSPICE simulation
 To explore the practicability aspect of the proposed circuit, the experimental realization using commercially available ICs is also explored and included
 
 GRAPHICAL ABSTRACT

Engineering negative coupling and corner modes in a three-dimensional acoustic topological network

Higher-order topological insulators have aroused massive attention due to their fancy topological properties. Artificial metamaterials, with their exquisite fabrications and measurements, offer an ideal approach to investigate topological properties in classical systems. Therein lower-dimensional corner states obtained in the three-dimensional (3D) Su-Schrieffer-Heeger (SSH) model and the octupole insulator have been extensively explored in acoustic systems. However, the existing acoustic negative couplings in the quantized multipole topological phases are mostly formed by shifting the connecting waveguides between every two resonators, which support a restricted dipole resonance mode. Here we establish subwavelength acoustic third-order topological insulators based on the theoretical frame of acoustic transmission lines and provide an innovative way for freely engineering acoustic negative couplings that work in arbitrary resonance modes. Furthermore, topological invariant calculations and numerical simulations in analogous acoustic networks validate the occurrence of topological corner states in both the 3D SSH model and octupole insulator. We foresee that the proposed methodology will broaden future avenues for studying atypical topological quantum effects with negative couplings in the classical macroscale context.

Stabilizing MXene-based nanofiltration membrane by forming analogous semi-interpenetrating network architecture using flexible poly(acrylic acid) for effective wastewater treatment

Two-dimension MXene (Ti3C2Tx) membrane has attracted widespread interest in water treatment to relieve the problem of clean water shortage. Nevertheless, notorious swelling and subsequent instability problems of the MXene-based membrane when immersing in water hinder its wide application, owing to the weak interaction between the adjacent MXene laminates. Herein, a feasible strategy is proposed to stabilize and tune the nanochannels of MXene-based membranes and achieve superior nanofiltration efficiency. Poly(acrylic acid) (PAA) macromolecules are firstly inserted into the MXene nanosheets and form PAA-modified MXene (PAA-MXene) dispersion. Then, the dispersion is vacuum filtered on the electrospun polyacrylonitrile (PAN) nanofibrous substrate to form an ultra-thin PAA-MXene hydrogel barrier layer, thus assembling into PAA-MXene/PAN composite membrane. The MXene laminates will be intertwined by the long chain of flexible PAA macromolecules and thus form homologous semi-interpenetrating network structure with MXene as physical crosslinker through strong hydrogen-bond interaction, which will restrict the obvious structural change of MXene nanosheets and effectively adjust their d-spacing at ∼14.28 Å with MXene: PAA mass ratio of 1:0.5. The synergic effect of PAA and MXene nanosheets, as well as electrospun PAN nanofibrous substrate, endows the PAA-MXene/PAN membrane with superior nanofiltration performance to various dyes. The optimum PAA-MXene/PAN membrane possesses permeate fluxes of 271.26, 516.34, 300.83 L m−2 h−1 and rejections of 98.92%, 99.52%, 98.12% to DR80, AB8GX and AB90 at 0.1 MPa, respectively, which exhibit distinct advantage than the MXene-based membrane in literature. In addition, the PAA-MXene/PAN membrane has good long-time durability and is more suitable for low pressure nanofiltration. All these make PAA-MXene/PAN membrane a promising nanofiltration membrane for wastewater treatment.

Mobile Two-Way Molecular Communication via Diffusion Using Amplify-and-Forward and Analog Network Coding.

Mobile molecular communication via diffusion (MCvD) has attracted lots of attentions due to its time-varying channels. In this paper, we investigate a mobile two-way MCvD model, which consists of two mobile source nanomachines and a mobile relay nanomachine. The amplify-and-forward (AF) relaying and analog network coding (ANC) are utilized to implement the exchange of information between two source nanomachines in this model. To explore the performance of the mobile two-way MCvD system, we first adopt the depleted molecule shift keying (D-MoSK) modulation, and then the mathematical expressions of symbol error probability (SEP) and mutual information are derived by using AF and ANC scheme on the basis of maximum a posteriori (MAP) probability detection. Finally, we present the numerical and simulation results. Compared with the AF-No-ANC scheme which is without use of ANC scheme, the scheme of combining AF and ANC can significantly improve the performance of SEP and mutual information. Moreover, the D-MoSK modulation outperforms the molecule shift keying (MoSK) modulation for this mobile two-way MCvD system. In addition, we propose the evaluation and discussion about the impacts of several important parameters on the performance of this system. These results can be used to design mobile two-way MCvD system with lower SEP and higher mutual information.

Multilayer patent citation networks: A comprehensive analytical framework for studying explicit technological relationships

The use of patent citation networks as research tools is becoming increasingly commonplace in the field of innovation studies. However, these networks rarely consider the contexts in which these citations are generated and are generally restricted to a single jurisdiction. Here, we propose and explore the use of a multilayer network framework that can naturally incorporate citation metadata and stretch across jurisdictions, allowing for a complete view of the global technological landscape that is accessible through patent data. Taking a conservative approach that links citation network layers through triadic patent families, we first observe that these layers contain complementary, rather than redundant, information about technological relationships. To probe the nature of this complementarity, we extract network communities from both the multilayer network and analogous single-layer networks, then directly compare their technological composition with established technological similarity networks. We find that while technologies are more splintered across communities in the multilayer case, the extracted communities match much more closely the established networks. We conclude that by capturing citation context, a multilayer representation of patent citation networks is, conceptually and empirically, better able to capture the significant nuance that exists in real technological relationships when compared to traditional, single-layer approaches. We suggest future avenues of research that take advantage of novel computational tools designed for use with multilayer networks.

Demonstration of a Hybrid Analog–Digital Transport System Architecture for 5G and Beyond Networks

In future mobile networks, the evolution of optical transport architectures enabling the flexible, scalable interconnection of Baseband Units (BBUs) and Radio Units (RUs) with heterogeneous interfaces is a significant issue. In this paper, we propose a multi-technology hybrid transport architecture that comprises both analog and digital-Radio over Fiber (RoF) mobile network segments relying on a dynamically reconfigurable optical switching node. As a step forward, the integration of the discussed network layout into an existing mobile infrastructure is demonstrated, enabling the support of real-world services through both standard digital and Analog–Intermediate- Frequency over Fiber (A-IFoF)-based converged fiber–wireless paths. Emphasis has been placed on the implementation of a real-time A-IFoF transceiver that is employed through a single embedded fully programmable gateway array (FPGA)-based platform that serves as an Ethernet to Intermediate Frequency (IF) bridge for the transmission of legacy traffic over the analog network segment. The experimental evaluation of the proposed concept was based on the dynamic optical routing of the legacy Common Public Radio Interface (CPRI), 1.5 GBaud analog-intermediate frequency-over-fiber (A-IFoF)/mmWave and 10 Gbps binary optical waveforms, showing acceptable error vector magnitude (EVM) values for the complex radio waveforms and error-free operation for binary optical streams, with Bit Error Rate (BER) values less than 10−9. Finally, the end-to-end proof-of-concept demonstration of the proposed solution was achieved through the delivery of 4K video streaming and Internet Protocol (IP) calls over a mobile core network.

Not every sample is efficient: Analogical generative adversarial network for unpaired image-to-image translation

Image translation is to learn an effective mapping function that aims to convert an image from a source domain to another target domain. With the proposal and further developments of generative adversarial networks (GANs), the generative models have achieved great breakthroughs. The image-to-image (I2I) translation methods can mainly fall into two categories: Paired and Unpaired. The former paired methods usually require a large amount of input–output sample pairs to perform one-side image translation, which heavily limits its practicability. To address the lack of the paired samples, CycleGAN and its extensions utilize the cycle-consistency loss to provide an elegant and generic solution to perform the unpaired I2I translation between two domains based on unpaired data. This thread of dual learning-based methods usually adopts the random sampling strategy for optimizing and does not consider the content similarity between samples. However, not every sample is efficient and effective for the desired optimization and leads to optimal convergence. Inspired by analogical learning, which is to utilize the relationships and similarities between sample observations, we propose a novel generic metric-based sampling strategy to effectively select samples from different domains for training. Besides, we introduce a novel analogical adversarial loss to force the model to learn from the effective samples and alleviate the influence of the negative samples. Experimental results on various vision tasks have demonstrated the superior performance of the proposed method. The proposed method is also a generic framework that can be easily extended to other I2I translation methods and result in a performance gain.

Simulation of Low Power Self-Selective Memristive Neural Networks for in situ Digital and Analogue Artificial Neural Network Applications

Self-selective memory devices are considered promising candidates for suppressing the undesired sneak path currents that appear within crossbar memory structures and compromise their performance during the write and read operations. Along these lines, in this work we present forming free SiO <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$_{\mathbf{2}}$</tex-math></inline-formula> -based resistive devices with inherent self-selection and self-compliance properties. The devices can operate in dual mode, since they can perform volatile threshold switching and non-volatile bipolar resistive behavior by regulating the external voltage amplitude. Interestingly, the devices exhibit low operating voltage ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\sim$</tex-math></inline-formula> 260 – 360 mV) and quick response time ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\sim$</tex-math></inline-formula> 100 ns). A comprehensive model is then developed in order to interpret this unique feature and provide valuable insights into the underlying physical mechanisms. Additionally, self-activated neural networks are theoretically investigated by performing in-situ digital and analog computations, whereas the calculated outcomes are compared with traditional neural networks with transistors as selecting elements. More specifically, a logic NAND gate and supervised learning upon the MNIST dataset is performed, while the proposed neural network architecture is tuned differently according to the application. Notably, the acquired results are comparable with the respective data where selectors have been employed, indicating the promising aspects of the proposed memristive devices. Moreover, a thorough analysis is carried out regarding the correlation between the device's linearity and the recognition's accuracy score, offering valuable insights. The proposed architecture paves the way for the development of energy-efficient artificial neural network computing architectures with tunable properties.

An Efficient Beamforming Architecture to Handle the Trade-Off Between Performance and Hardware Complexity in Multiuser Massive MISO Systems

Beamforming is the fundamental concept of wireless communications to serve several users through multiple-antenna transceivers. The advent of massive multiple-input multiple-output (MIMO) systems leads to an investigation into beamforming to reach optimal performance. However, fully-digital beamforming implementation has to face important challenges in terms of consumed power and cost of the RF chains and converters. To solve the problem, some studies focus on reducing hardware complexity by dividing the beamforming into digital and analog parts, known as hybrid beamforming (HBF). In HBF systems, the dominant part of the hardware complexity depends on the architecture of the analog network, which determines the required RF components. In this paper, we categorize the analog beamforming parts based on the connectivity between RF-chains and antennas. Moreover, we show the impacts of the connection on the analog beamforming matrix. To this aim, we propose an efficient connectivity architecture in which the antennas are divided into fully-connected and singly-connected groups so that we can deal with the hardware complexity-performance trade-off. In addition, to achieve further performance improvements, we propose a dynamic architecture that adjusts the connection of the RF-chain antenna to the channel state information (CSI) through a switch network. Analytically, we propose an efficient approach to calculate the zero-forcing precoder, which is proper for both fixed and dynamic architectures. A two-part algorithm based on the greedy search method has also been developed to obtain switch states in dynamic architecture and then implemented by a deep neural network (DNN). The simulation results confirm the theoretical analysis and the suitability of the proposed architecture.

Demonstrating Analog Inference on the BrainScaleS-2 Mobile System

We present the BrainScaleS-2 mobile system as a compact analog inference engine based on the BrainScaleS-2 ASIC and demonstrate its capabilities at classifying a medical electrocardiogram dataset. The analog network core of the ASIC is utilized to perform the multiply-accumulate operations of a convolutional deep neural network. At a system power consumption of 5.6W, we measure a total energy consumption of 192uJ for the ASIC and achieve a classification time of 276us per electrocardiographic patient sample. Patients with atrial fibrillation are correctly identified with a detection rate of (93.7${\pm}$0.7)% at (14.0${\pm}$1.0)% false positives. The system is directly applicable to edge inference applications due to its small size, power envelope, and flexible I/O capabilities. It has enabled the BrainScaleS-2 ASIC to be operated reliably outside a specialized lab setting. In future applications, the system allows for a combination of conventional machine learning layers with online learning in spiking neural networks on a single neuromorphic platform.

Drone Assisted Network Coded Cooperation With Energy Harvesting: Strengthening the Lifespan of the Wireless Networks

Next generation wireless systems include battery operated devices which demand higher throughput and a better reliability in an energy efficient fashion. To fulfil these requirements, in this paper, we propose a novel scenario where we include a dynamic Wireless Power Splitting (WPS) factor for Energy Harvesting (EH) at nodes in a Drone Assisted Network Coded Cooperation (DA-NCC) system. The dynamic WPS factor used for EH in DA-NCC system is made more realistic by determining through the probability of Line-of-Sight (LoS) occurrence. Analytical framework is developed for residual Analog Network Coding (ANC) noise and variance of ANC-noise in EH scenario. We also derive the average rate and average outage probability expressions for the proposed channel model. Various algorithms are developed for deciding the Air-to-Ground (A2G) channel distributions, harvesting the energy at relay and source nodes and evaluating the performance metrics of our proposed work. Our investigations reveal that the use of EH in DA-NCC improves the lifespan of the network. Our findings play important roles in disaster management scenarios where cellular connections to base stations are disrupted due to natural calamities and battery constrained drones are deployed for assistance.

Enhanced iodine capture by incorporating anionic phosphate unit into porous networks

Although significant advance has been achieved in developing porous organic polymers for effective iodine capture, the underlying principle for the adsorption behavior is not fully understood. Herein, analogous anionic porous networks (IPN-CSUs) bearing phosphate unit were fabricated via a one-pot Friedel-Crafts alkylation, in which heteroatom-containing triphenylamine (N) and 9,10-anthraquinone (O) were picked up as co-building blocks. The compositions were simply regulated, and the obtained networks displayed rich porous structure, thereby offering high sorption capability towards iodine in cyclohexane solution. Specifically, the anionic networks exhibited an improvement up to 122.4% in capturing iodine compared to the corresponding neutral counterpart. The adsorption performance of IPN-CSU-1 (N incorporated) in iodine vapor reached 357 wt%, which is comparable to the state-of-art networks possessing ultra-high specific surface area and the performance maintained at about 82% of the original value after five consecutive cycles. In addition, the combination of experimental and theoretical computational simulations highlights the prominent role of negatively charged phosphate group in efficient iodine capture.

Low-cost lumped parameter modelling of hydrogen storage in solid-state materials

A low-cost lumped parameter model (LPM) is developed to simulate hydrogen storage in solid-state materials. The proposed LPM is a zero-dimensional model based on an analogous transient thermal resistance network that considers the whole storage system as a resistor–capacitor (RC) circuit with a current source. Combined with the mass conservation, sorption kinetics, and equation of state, the LPM is capable of predicting key thermofluidic quantities of the storage system, such as the storage bed temperature and pressure, hydrogen fraction and flowrate, and storage tank temperature. To test the applicability of the LPM, chemisorptive (Mg and LaNi5) and physisorptive (activated carbon) solid-state materials are simulated for adsorption and desorption processes. For both material types, LPM achieves predictive accuracies comparable to three-dimensional computational fluid dynamics (CFD) simulations, especially for hydrogen fraction and storage times (with maximum errors less than 9.6%), for a fraction of the computational cost (23,734 times lower memory requirement and 126 times faster calculation time). In addition, several formulations of the bed thermal resistance are proposed and discussed in terms of their impact on the accuracy of the LPM predictions. For slender cylindrical tanks, an equivalent annulus formula achieves the best balance of simplicity and accuracy.

A Cluster of FPAAs to Recognize Images Using Neural Networks

Analog computing has been recovering its relevance in the recent years. Field-Programmable Analog Arrays (FPAAs) are the equivalent to Field-Programmable Gate Arrays (FPGAs) but in the analog and mixed-signal domain. In order to increase the amount of analog resources, in this brief a cluster of 40 FPAAs is proposed. As a use case, a 19-8-6-4 feedforward Neural Network has been implemented on such cluster. With the help of a DCT-based software framework, this NN is able to classify <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$28 \times 28$ </tex-math></inline-formula> images from MNIST. Results show that the analog network is able to obtain similar results as the software baseline network.

Replacement of In-Orbit Spacecraft Attitude Determination Algorithms with Adaptive Neuro-Fuzzy Inference System via Subtractive Clustering

Traditional spacecraft attitude determination algorithms suffer from accuracy versus computational load dilemma. As algorithm complexity increases, its computational load increases and also its accuracy. This limits algorithm applicability in real-time (In-Orbit) environment. In this article, a solution based on Adaptive Neuro-Fuzzy Inference System (ANFIS) has been developed. The number of inputs to ANFIS is 12, which implies an extremely large number of rules of the membership functions to be utilized. Consequently, the proposed solution blows up. Hence, a clustering technique should be utilized in order to minimize the number of rules. Design simulations of the former Egyptian spacecraft EGYPTSAT-1 have been utilized as test case. During the standby operational mode, the spacecraft is assumed to have attitude errors within ±10°. ANFIS is a multi-input single-output algorithm. Therefore, a separate ANFIS is developed for each one of the attitude angles. Each one of them is trained based on two data sets, provided by TRIAD and Q-method simulators. Thus, six ANFISs are developed. The results have shown that the developed ANFISs can work in the stand-by operational mode and provide attitude estimates of ±10°. Moreover, the average execution time for ANFIS is about 40% of the average execution time for the analogous Cascade-Forward Neural Network algorithms (CFNN) and 25% of the Extended Kalman Filter average execution time. The computational overhead of the proposed ANFIS algorithms is constant regardless of the attitude determination or estimation algorithm to be replaced by ANFIS. Consequently, very accurate and sophisticated models could be built in order to provide training data set and then replaced by ANFIS without increasing the ANFIS computational load.

Fluvial Features on Titan and Earth: Lessons from Planform Images in Low-resolution SAR

Abstract Cassini Synthetic Aperture Radar (SAR) images of Titan’s surface revealed river networks strikingly similar to those on Earth. However, Cassini SAR has low spatial resolution and image quality compared to data used to map channels on Earth, so traditional methods for characterizing river networks might not yield accurate results on Titan. We mapped terrestrial analog networks with varying resolutions and image qualities to determine which geomorphologic metrics were invariant with scale or resolution. We found that branching angle and drainage density varied significantly with image resolution, and we therefore expect the actual drainage density of Titan’s channel networks to be significantly higher than the values calculated from Cassini data. Calculated network geometry did not change predictably with resolution and would therefore not be an ideal metric for interpreting Titan’s channel networks. The measured channel width, basin length and width, and drainage area all behaved predictably as resolution varied, leading us to conclude that these metrics could be applied to Cassini data. We then mapped all observable fluvial features on Titan—excluding those in the highly incised labyrinth terrains—visible in the Cassini data set. In our new maps, we observe differences in basin shapes between Titan’s polar and equatorial regions and dichotomies in the relative channel density between the northern and southern midlatitudes and poles of Titan: channels are concentrated at the poles and southern midlatitudes. These patterns may reflect differences in bedrock material and/or different climate histories.

The city as a theatre of characters. John Hejduk’s Masques

The architectural work of John Hejduk (1929-2000) is marked by theoretical-design research, collected in series with titles and periods. Among these series the one entitled Masques, developed since about 1979, can be considered the nucleus of his research on the architecture of the city and the place of origin of his language of construction later developed in his realized buildings. This paper analyses the dense network of references and analogies established by Hejduk to create his Masques, trying to fix its origin in the idea of the city as a theatre of characters composed of architecture. Starting from the name chosen for the title of this series, the paper tries to trace the threads that lead from the general work of the various projects of the Masques series, to the reflections and ideas that produced it. Then, it comes back again to the observation of architecture and of a case study (Security, 1989), to understand and explain its meaning and the compositional methods involved into the process of genesis of form. Through the entire work named Masques, and its recognizable link with the buildings and installation realized around the world, Hejduk has built an archive of architectural prototypes ready to construct different parts of the city, thus highlighting the strong connection that his work establishes with reality in order "to conceive it, represent it and finally realize it".

Reprocessable polylactide-based networks containing urethane and disulfide linkages

One of the ways of improving the mechanical properties of polylactide (PLA), which is produced from renewable resources, is crosslinking. In this contribution crosslinked polylactide was obtained by coupling of PLA star polymers using aliphatic diisocyanate, resulting in poly(ester-urethanes). To obtain crosslinked material, able to reprocess, disulfide groups undergoing exchange reactions were introduced into the PLA network structure during the synthesis step. Networks without disulfide groups both, without and with additional low molecular weight diol (diethylene glycol), replacing diol containing disulfide bond (2-hydroxyethyl disulfide) were also prepared for comparative study of mechanical and rheological properties. DSC and FT IR analyses of the networks indicated the presence of hydrogen bonding in obtained materials which influences the ability of disulfide linkages to exchange. The networks with disulfide groups were able to rearrange only at a temperature at which hydrogen bonds were broken (above 100 °C). The ability to rearrange network structure manifested itself in lower storage modulus (G’) above glass transition (Tg = 38–48 °C) as well as faster stress relaxation in comparison with the analogous network without disulfide groups. A PLA-based network containing a sufficiently high density of disulfide groups could be reprocessed at a temperature above 170 °C.