Ericsson (Sweden)

The landscape of modern computing systems is shifting towards architectures built by combining available services under the “everything as a service” paradigm. These architectures are deployed on distributed cloud-edge infrastructures, aiming to provide innovative services to a wide range of users. However, it is crucial for these systems to address environmental sustainability concerns. This poses challenges in operating such systems in open, dynamic, and uncertain environments while minimizing their energy consumption. To tackle these challenges, we propose a decentralized service assembly approach that ensures the assembly is energetically self-sustainable by relying on locally harvested and stored energy. In our contribution, we introduce a general service selection template that enables the derivation of different selection policies. These policies guide the construction and maintenance of the service assembly. To evaluate their effectiveness in meeting the sustainability requirements, we conduct a comprehensive set of simulation experiments, providing valuable insights.

The transition from hospital to home after stroke or transient ischaemic attack (TIA) is a vulnerable phase marked by gaps in coordination, communication and patient preparedness. This study evaluated the effectiveness of a multicomponent care transition support on perceived quality of care transition (QCT) amongst people with stroke or TIA one week post-discharge. This non-randomised controlled study was conducted in three Swedish hospitals. Adult people with stroke or TIA discharged to home and referred for home-based neurorehabilitation were eligible. Data were collected during hospitalisation and at a blinded one-week follow-up. Primary outcome was perceived QCT, assessed with the Care Transition Measure. Secondary outcomes included health literacy, medication adherence and perception of received care. In total, 163 participants completed follow-up. The intervention group reported higher perceived QCT than controls (p = 0.002), which remained significant after adjustments (β = 6.34; 95% CI 0.56–12.12; p = 0.03). The intervention group reported greater confidence in managing health and better understanding of medications. Health literacy was higher in the intervention group, with no between-group differences in medication adherence or perception of received care. A multicomponent care transition support intervention may improve perceived QCT early after discharge in people with stroke or TIA.

Large Language Models (LLMs) are increasingly integrated into telecommunications for customer service, network analytics, and operational support. While offering efficiency and innovation, their deployment raises critical privacy, security, and ethical challenges due to the uniquely sensitive and longitudinal nature of telecom data. This paper presents a survey of LLM risks and safeguards in telecom context. Our survey introduces a telecom-specific taxonomy linking privacy, security, and ethics with use cases and outlines open research challenges such as longitudinal data privacy, cross-border compliance, and scalable guardrails.

Artificial Neural Networks (ANNs) have become foundational to modern intelligent systems, powering applications ranging from cloud infrastructures to low-power edge devices. Spiking Neural Networks (SNNs), with an additional promise of low energy compute, offer an attractive alternative for deployment in domains where energy efficiency is of paramount importance. In this work, we address the problem of output safety verification for SNNs, wherein we aim to identify inputs that lead the network to violate specified safe output ranges. We propose a novel two-way verification framework tailored for SNNs employing Leaky Integrate-and-Fire (LIF) neuron models. Our method integrates temporal encoding with scalable constraint solving to effectively explore the input space. Experimental evaluation on benchmarks demonstrates that our approach achieves superior performance on larger networks compared to existing techniques.

A major challenge hampering the industrial and biological exploitation of lipases is the self-limiting effect of the lipolytic reaction. Defying sustained attention throughout history, the influence of lipase catalysed lipolysis on the internal structure of thin triglyceride films has proved elusive. An in-depth understanding of the lipolysis process at the triglycerides/aqueous interface will assist in creating innovative methods to enhance yields. This study furthers our understanding of the influence of solution pD (pH equivalent for aqueous solutions prepared with D2O) on the effect of lipolysis on the structure of thin triolein films. All experiments were performed at pD 7.0 or pD 8.5, either side of the apparent pKa of the primary product of the hydrolysis (oleic acid/oleate). Spectroscopic ellipsometry measurements were employed to kinetically track the changes in thickness of the triolein film after the introduction of 2 ppm Thermomyces lanuginosus lipase (TLL). Neutron reflectometry experiments revealed the internal structure of the thin films before and after TLL digestion, while fast kinetic measurements capturing changes to the reflectivity profile throughout lipolysis. Both techniques revealed significant variations in the physical properties and enzymatic conversion of the triolein films between pD 7.0 and pD 8.5.

Objectives(s):To report methodology that has been developed to provide real-time monitoring of radiation emissions from subjects treated with radionuclide therapies and summarise the radioiodine retention profiles of 250 subjects treated for differentiated thyroid cancer with 131I.Methods: A small ceiling-mounted radiation detector for continuously monitoring the exposure rate in the radiation isolation rooms has been developed. Measurements were made every minute after administration of 1-6 GBq of 131I over the one to three days typical inpatient admission. The data are saved in text format and have been fitted with a mono-exponential curve to measure retention half time.Results:The average effective retention half time (t½ (Eff)) for all subjects was 11.9±3.2 hrs (range: 5.0–23.1 hrs; n=250). Over 90% of the subjects had their serum TSH levels increased by injection of recombinant human TSH prior to treatment. Average retention half-time was found to be less in subjects lower than 55 year of age (t½ (Eff)=11.5 hrs) compared to those 55 or older (t½ (Eff)=14.4 hrs) (P=0.0007). Conclusions: Despite the subjects, being free to move around the isolation room during admission and thus changing the source-detector geometry markedly, the system has been able to characterise their retention profiles after radioiodine treatment. These real-time measurements have applications in planning therapy and monitoring the subjects during their admission to the hospital and can be used for “live” updates for all staff as well as providing insights into the fate of radioiodine in the body.

There is a growing interest in equipping cellular devices with satellite connectivity and the 3rd Generation Partnership Project (3GPP) continues to further evolve 5G advanced non-terrestrial networks (NTN) technology since its inception in Release-17 (Rel-17). In Rel-19, 3GPP introduced further enhancements for New Radio (NR) NTN as well as Narrowband Internet of Things (NB-IoT) NTN. In this paper, a technical overview of the Rel-19 physical layer enhancements for NR NTN and NB-IoT NTN is provided. The design details are described for the following features of Rel-19 NR NTN: 1) enhanced downlink coverage; 2) improved uplink (system) capacity; and 3) support of half duplex (HD) frequency-division duplex (FDD) operation for Reduced Capability (Red-Cap) and enhanced RedCap devices. Similarly, for NB-IoT NTN: 1) the uplink (system) capacity enhancements and 2) the support of time-division duplex operation are summarized. Finally, evaluation results are presented for Rel-19 NTN features such as downlink coverage enhancements and uplink capacity enhancements.

Abstract In this paper, a new framework – RoadRAT - for analyzing the probability of inundation, wave runup, and erosion is applied within the context of risk management for coastal roads. The analysis focus on a case study in south Sweden and consists of risk identification, calculation of probability, and discussion of consequences and mitigation methods. The results show that future risks are highly dependent on the future greenhouse gas emissions.