IMDEA Networks Institute Digital Repository
Not a member yet
1915 research outputs found
Sort by
A Snapshot of COVID-19 Incidence, Hospitalizations, and Mortality from Indirect Survey Data in China in January 2023 (Extended Abstract)
Get PDFThe estimation of incidence has been a crucial component for monitoring COVID-19 dissemination. This has become challenging when official data are unavailable or insufficiently reliable. Hence, the implementation of efficient, inexpensive, and secure techniques that capture information about epidemic indicators is required. This study aims to provide a snapshot of COVID-19 incidence, hospitalizations, and mortality in different countries in January 2023. To this end, we collected data on the number of cases, deaths, vaccinations, and hospitalizations among the fifteen closest contacts. More precisely, indirect surveys were conducted for 100 respondents from Australia on 19 January 2023, 200 respondents from the UK on 19 January 2023, and 1,000 respondents from China between 18-26 January 2023. To assess the incidence of COVID-19, we used a modified version Network Scale-up Method (NSUM) that fixes the number of people in the contact network (reach). We have compared our estimates with official data from Australia and the UK in order to validate our approach. In the case of the vaccination rate, our approach estimates a very close value to the official data, and in the case of hospitalizations and deaths, the official results are within the confidence interval. Regarding the remaining variables, our approach overestimates the values obtained by the Our World in Data (OWID) platform but is close to the values provided by the Officer of National Statistics (ONS) in the case of the UK (within the confidence interval). In addition, Cronbach's alpha gives values that allow us to conclude that the reliability of the estimates in relation to the consistency of the answers is excellent for the UK and good for Australia. Following the same methodology, we have estimated the same metrics for different Chinese cities and provinces. It is worth noting that this approach allows quick estimates to be made with a reduced number of surveys to achieve a wide population coverage, preserving the privacy of the participants.Spanish Ministry of Science and InnovationEuropean Union "Next Generation EU"TRUEinpres
Online Algorithms for Hierarchical Inference in Deep Learning applications at the Edge
Get PDFWe consider a resource-constrained Edge Device (ED) embedded with a small-size ML model (S-ML) for a generic classification
application, and an Edge Server (ES) that hosts a large-size ML model (L-ML). Since the inference accuracy of S-ML is lower than that
of the L-ML, offloading all the data samples to the ES results in high inference accuracy, but it defeats the purpose of embedding S-ML
on the ED and deprives the benefits of reduced latency, bandwidth savings, and energy efficiency of doing local inference. To get the
best out of both worlds, i.e., the benefits of doing inference on the ED and the benefits of doing inference on ES, we explore the idea
of Hierarchical Inference (HI), wherein S-ML inference is only accepted when it is correct, otherwise the data sample is offloaded for
L-ML inference. However, the ideal implementation of HI is infeasible as the correctness of the S-ML inference is not known to the ED.
We thus propose an online meta-learning framework to predict the correctness of the S-ML inference. The resulting online learning
problem turns out to be a Prediction with Expert Advice (PEA) problem with continuous expert space. We consider the full feedback
scenario, where the ED receives feedback on the correctness of the S-ML once it accepts the inference, and the no-local feedback
scenario, where the ED does not receive the ground truth for the classification, and propose the HIL-F and HIL-N algorithms and
prove a regret bound that is sublinear with the number of data samples.We evaluate and benchmark the performance of the proposed
algorithms for image classification applications using four datasets, namely, Imagenette and Imagewoof [18],MNIST [24], and CIFAR-
10 [23].E
A generic architecture of an affective recommender system for e-learning environments
Get PDFPersonalization of suggestions of contents plays a key role in adaptive virtual learning environments. Good recommendations can raise the interest of students in the learning process, while, on the other hand, bad recommendations can have catastrophic results for the learning process. The affective state of the student is a very influential factor in the learning process. In this work, a generic architecture of an affective recommender system for e-learning environments is developed, to serve as a guide for future implementations of this kind of recommender system. Here, the affective characteristics of students are represented by their personalities, learning styles, emotional states, and expertise levels. Thus, the main contribution is the proposition of a generic architecture of an affective recommendation system for the educational field. The architecture is completely modular, which gives it great flexibility because the emotion engine is separated from the personal characteristics engine, and is not based on specific models of emotions. This work finishes with examples of use cases of the architecture. According to the results in these examples, our architecture is capable of analyzing the polarity of academic documents based on their content, determining the personal characteristics of students (including their emotions), and from there, recommending learning resources to students considering emotions as the main element of the process.TRUEpu
waveSLAM: Empowering accurate indoor mapping using off-the-shelf millimeter-wave self-sensing
Get PDFThis paper presents the design, implementation and evaluation of waveSLAM, a low-cost mobile robot system that uses the millimetre wave (mmWave) communication devices to enhance the indoor mapping process targeting environments with reduced visibility or glass/mirror walls. A unique feature of waveSLAM is that it only leverages existing Commercial-Off-The-Shelf (COTS) hardware (Lidar and mmWave radios) that are mounted on mobile robots to improve the accurate indoor mapping achieved with optical sensors. The key intuition behind the waveSLAM design is that while the mobile robots moves freely, the mmWave radios can periodically exchange angle and distance estimates between themselves (self-sensing) by bouncing the signal from the environment, thus enabling accurate estimates of the target object/material surface. Our experiments verify that waveSLAM can archive cm-level accuracy with errors below 22 cm and 20◦ in angle orientation which is compatible with Lidar when building indoor maps.TRUEinpres
Scalable multi-modal learning for cross-link channel prediction in massive IoT networks
Get PDFTomorrow's massive-scale IoT sensor networks are poised to drive uplink traffic demand, especially in areas of dense deployment. To meet this demand, however, network designers leverage tools that often require accurate estimates of Channel State Information (CSI), which incurs a high overhead and thus reduces network throughput. Furthermore, the overhead generally scales with the number of clients, and so is of special concern in such massive IoT sensor networks. While prior work has used transmissions over one frequency band to predict the channel of another frequency band on the same link, this paper takes the next step in the effort to reduce CSI overhead: predict the CSI of a nearby but distinct link. We propose Cross-Link Channel Prediction (CLCP), a technique that leverages multi-view representation learning to predict the channel response of a large number of users, thereby reducing channel estimation overhead further than previously possible. CLCP's design is highly practical, exploiting existing transmissions rather than dedicated channel sounding or extra pilot signals. We have implemented CLCP for two different Wi-Fi versions, namely 802.11n and 802.11ax, the latter being the leading candidate for future IoT networks. We evaluate CLCP in two large-scale indoor scenarios involving both line-of-sight and non-line-of-sight transmissions with up to 144 different 802.11ax users and four different channel bandwidths, from 20 MHz up to 160 MHz. Our results show that CLCP provides a 2× throughput gain over baseline and a 30% throughput gain over existing prediction algorithms.TRUEinpres
The Case for Hierarchical Deep Learning Inference at the Network Edge
Get PDFResource-constrained Edge Devices (EDs), e.g., IoT sensors and microcontroller units, are expected to make intelligent decisions using Deep Learning (DL) inference at the edge of the network. Toward this end, developing tinyML models is an area of active research - DL models with reduced computation and memory storage requirements - that can be embedded on these devices. However, tinyML models have lower inference accuracy. On a different front, DNN partitioning and inference offloading techniques were studied for distributed DL inference between EDs and Edge Servers (ESs). In this paper, we explore Hierarchical Inference (HI), a novel approach proposed in [19] for performing distributed DL inference at the edge. Under HI, for each data sample, an ED first uses a local algorithm (e.g., a tinyML model) for inference. Depending on the application, if the inference provided by the local algorithm is incorrect or further assistance is required from large DL models on edge or cloud, only then the ED offloads the data sample. At the outset, HI seems infeasible as the ED, in general, cannot know if the local inference is sufficient or not. Nevertheless, we present the feasibility of implementing HI for image classification applications. We demonstrate its benefits using quantitative analysis and show that HI provides a better trade-off between offloading cost, throughput, and inference accuracy compared to alternate approaches.TRUEpu
A Framework for Wireless Technology Classification using Crowdsensing Platforms
Get PDFSpectrum crowdsensing systems do not provide labeled data near real-time yet. We propose a framework that addresses this challenge and relies solely on Power Spectrum Density (PSD) data collected by low-cost receivers. A major hurdle is to design a system that is computationally efficient for near real-time operation, yet using only the limited 2 MHz bandwidth of low-cost spectrum sensors. First, we present a method for unsupervised transmission detection that works with PSD data already collected by the backend of the crowdsensing platform, and that provides stable detection of transmission boundaries. Second, we introduce a data-driven deep learning solution to classify the wireless technology used by the transmitter, using transmission features in a compressed space extracted from single PSD measurements over at most 2 MHz band. We build an experimental platform, and evaluate our framework with real-world data collected from 47 different sensors deployed across Europe. We show that our framework yields an average classification accuracy close to 94.25% over the testing dataset, with a maximum latency of 3.4 seconds when integrated in the backend of a major crowdsensing network. Code and data have been released for reproducibility and further studies.TRUEinpres
Full Network Sensing: Architecting 6G beyond Communications
Get PDFAs requirements for 6G systems start being shaped, we propose the provoking vision that the future generations of mobile networks shall reach beyond their traditional role of enablers for communication services. In fact, mobile networks have an impressive untapped potential for dual use as a pervasive remote sensing platform, which, if duly exploited, can support compelling original applications across a variety of domains. Unlocking this capacity requires a systematic end-to-end integration of sensing functionalities in the mobile network architecture beyond radio access, on a level and in synergy with classical data transport operations. In this paper, we formalize the vision above as the full network sensing paradigm. We discuss representative use cases for full network sensing, and requirements towards its implementation. We lay out a conceptual model for the integration and management of sensing functionalities into modern mobile network architectures, and present open technical challenges to realizing the full network sensing vision. These contributions lay the very first principles to a disruptive paradigm that can unlock new markets for mobile network providers, and support innovative remote sensing services that benefit the society at large.Regional Government of MadridTRUEpu
Digital Twins for Next-Generation Mobile Networks: Applications and Solutions
Get PDFDigital Twins (DTs) create fully-synchronized virtual representations of real-world systems, which can serve as interactive counterparts for artificial intelligence (AI) and machine learning (ML) algorithms, and hold significant importance for the upcoming 6G mobile networks. In this paper, we argue that DTs can improve all phases of the intelligent networks' workflow, due to their adaptability and scalability properties that would allow them to transparently integrate new AI/ML algorithms faster, more scalably, and more precisely. Our contribution is two-fold: first, we propose three specific application scenarios of DT-enhanced network architectures in the context of 6G. Second, using open-source tools, we implement and evaluate in detail one of them. Our results demonstrate that our DT reflects the characteristics of the physical object, successfully and scalably twinning it, and adapting to changing contextual conditions.TRUEpu