Academic literature on the topic 'Machine-to-Machine, Device-to-Device, Resource allocation'

In Long Term Evolution-Advanced (LTE-A), network-controlled device-to-device (D2D) communications allow User Equipments (UEs) to communicate directly, without involving the Evolved Node-B in data relaying, while the latter still retains control of resource allocation. The above paradigm allows reduced latencies for the UEs and increased resource efficiency for the network operator, and is therefore foreseen to support several services, from Machine-to-machine to vehicular communications. D2D communications introduce research challenges that might affect the performance of applications and upper-layer protocols, hence simulations represent a valuable tool for evaluating these aspects. However, simulating D2D features might pose additional computational burden to the simulation environment. To this aim, a careful modeling is required to reduce computational overhead. In this paper, we describe our modeling of network-controlled D2D communications in SimuLTE, a system-level LTE-A simulation library based on OMNeT++. We describe the core modeling choices of SimuLTE, and show how these allow an easy extension to D2D communications. Moreover, we describe in detail the modeling of specific problems arising with D2D communications, such as scheduling with frequency reuse, connection mode switching and broadcast transmission. We document the computational efficiency of our modeling choices, showing that simulation of D2D communications is not more complex than simulation of classical cellular communications of comparable scale. Results show that the heaviest computational burden of D2D communication lies in estimating the Sidelink channel quality. We show that SimuLTE allows one to evaluate the interplay between D2D communication and end-to-end performance of UDP- and TCP-based services. Moreover, we assess the accuracy of using a binary interference model for frequency reuse, and we evaluate the trade-off between speed of execution and accuracy in modeling the reception probability.

Internet of Things (IoT) is revolutionising technical environment of traditional methods as well as has applications in smart cities, smart industries, etc. Additionally, IoT enabled models' application areas are resource-constrained as well as demand quick answers, low latencies, and high bandwidth, all of which are outside of their capabilities. The above-mentioned issues are addressed by cloud computing (CC), which is viewed as a resource-rich solution. However, excessive latency of CC prevents it from being practical. The performance of IoT-based smart systems suffers from longer delay. CC is an affordable, emergent dispersed computing pattern that features extensive assembly of diverse autonomous methods. This research propose novel technique resource allocation and task scheduling for device-device communication in mobile Cloud IoT environment based on 5G networks. Here the resource allocation has been carried out using virtual machine based markov model infused wavelength division multiplexing. Task scheduling is carried out using meta-heuristic moath flame optimization with chaotic maps. So, by scheduling tasks in a smaller search space, system resources are conserved. We run simulation tests on benchmark issues and real-world situations to confirm the effectiveness of our suggested approach. The parameters measured here are resource utilization of 95%, response time of 89%, computational cost of 35%, power consumption of 38%, QoS of 85%.

As machine learning becomes ubiquitous, the need to deploy models on real-time, embedded systems will become increasingly critical. This is especially true for deep learning solutions, whose large models pose interesting challenges for target architectures at the “edge” that are resource-constrained. The realization of machine learning, and deep learning, is being driven by the availability of specialized hardware, such as system-on-chip solutions, which provide some alleviation of constraints. Equally important, however, are the operating systems that run on this hardware, and specifically the ability to leverage commercial real-time operating systems which, unlike general purpose operating systems such as Linux, can provide the low-latency, deterministic execution required for embedded, and potentially safety-critical, applications at the edge. Despite this, studies considering the integration of real-time operating systems, specialized hardware, and machine learning/deep learning algorithms remain limited. In particular, better mechanisms for real-time scheduling in the context of machine learning applications will prove to be critical as these technologies move to the edge. In order to address some of these challenges, we present a resource management framework designed to provide a dynamic on-device approach to the allocation and scheduling of limited resources in a real-time processing environment. These types of mechanisms are necessary to support the deterministic behavior required by the control components contained in the edge nodes. To validate the effectiveness of our approach, we applied rigorous schedulability analysis to a large set of randomly generated simulated task sets and then verified the most time critical applications, such as the control tasks which maintained low-latency deterministic behavior even during off-nominal conditions. The practicality of our scheduling framework was demonstrated by integrating it into a commercial real-time operating system (VxWorks) then running a typical deep learning image processing application to perform simple object detection. The results indicate that our proposed resource management framework can be leveraged to facilitate integration of machine learning algorithms with real-time operating systems and embedded platforms, including widely-used, industry-standard real-time operating systems.

Mobile networks have a great challenge by serving the expected billions of Internet of Things (IoT) devices in the upcoming years. Due to the limited simultaneous access in the mobile networks, the devices should compete between each other for resource allocation during a Random-Access procedure. This contention provokes a non-depreciable delay during the device’s registration because of the great number of collisions experienced. To overcome such a problem, a framework called Random-Access Accelerator (RAA) is proposed in this work, in order to speed up network access in massive Machine Type Communication (mMTC). RAA exploits Device-To-Device (D2D) communications, where devices with already assigned resources act like relays for the rest of devices trying to gain access in the network. The simulation results show an acceleration in the registration procedure of 99%, and a freed space of the allocated spectrum until 74% in comparison with the conventional Random-Access procedure. Besides, it preserves the same device’s energy consumption compared with legacy networks by using a custom version of Bluetooth as a wireless technology for D2D communications. The proposed framework can be taken into account for the standardization of mMTC in Fifth-Generation-New Radio (5G NR).

Massive machine-type communication (mMTC) in 5G New Radio (5G-NR) or the Internet of Things (IoT) is a network of physical devices such as vehicles, smart meters, sensors, and smart appliances, which can communicate and interact in real time without human intervention. In IoT systems, the number of networked devices is expected to be in the tens of billions, while radio resources remain scarce. To connect the massive number of devices with limited bandwidth, it is crucial to develop new access solutions that can improve resource efficiency and reduce control overhead as well as access delay. The key idea is controlling the number of arrival devices that want to access the system, and then allowing only the strongest device (that has the largest channel gain and each device is able to check whether it is the strongest device) be able to transit to BS. In this paper, we consider a random access problem in massive MIMO context for the collision resolution, in which the access class barring (ACB) factor is dynamically adjusted in each time slot to maximize access success rate for the strongest-user collision resolution (SUCRe) protocol. We propose the dynamic ACB scheme to find optimal ACB factor in the next time slot and then apply SUCRe protocol to achieve a good performance. This method is called dynamic access class barring combined strongest-user collision resolution (DACB-SUCR). In addition, we investigate two different ACB schemes that consist of the fixed ACB and the traffic-aware ACB to compare with the proposed dynamic ACB. Analysis and simulation results demonstrate that, compared with SUCRe protocol, the proposed DACB-SUCR method can remarkably reduce pilot collision, and increase access success rate. It is also shown that the dynamic ACB gives better performance than the fixed ACB and the traffic-aware ACB.

Recent advances in mobile technologies have facilitated the development of a new class of smart city and fifth-generation (5G) network applications. These applications have diverse requirements, such as low latencies, high data rates, significant amounts of computing and storage resources, and access to sensors and actuators. A heterogeneous private edge cloud system was proposed to address the requirements of these applications. The proposed heterogeneous private edge cloud system is characterized by a complex and dynamic multilayer network and computing infrastructure. Efficient management and utilization of this infrastructure may increase data rates and reduce data latency, data privacy risks, and traffic to the core Internet network. A novel intelligent middleware platform is proposed in the current study to manage and utilize heterogeneous private edge cloud infrastructure efficiently. The proposed platform aims to provide computing, data collection, and data storage services to support emerging resource-intensive and non-resource-intensive smart city and 5G network applications. It aims to leverage regression analysis and reinforcement learning methods to solve the problem of efficiently allocating heterogeneous resources to application tasks. This platform adopts parallel transmission techniques, dynamic interface allocation techniques, and machine learning-based algorithms in a dynamic multilayer network infrastructure to improve network and application performance. Moreover, it uses container and device virtualization technologies to address problems related to heterogeneous hardware and execution environments.

The recent Wi-Fi HaLow technology focuses on adopting Wi-Fi for the needs of the Internet of Things. A key feature of Wi-Fi HaLow is the Restricted Access Window (RAW) mechanism that allows an access point to divide the sensors into groups and to assign each group to an exclusively reserved time interval where only the stations of a particular group can transmit. In this work, we study how to optimally configure RAW in a scenario with a high number of energy harvesting sensor devices. For such a scenario, we consider a problem of device grouping and develop a model of data transmission, which takes into account the peculiarities of channel access and the fact that the devices can run out of energy within the allocated intervals. We show how to use the developed model in order to determine the optimal duration of RAW intervals and the optimal number of groups that provide the required probability of data delivery and minimize the amount of consumed channel resources. The numerical results show that the optimal RAW configuration can reduce the amount of consumed channel resources by almost 50%.

IEEE 802.11ah, known as Wi-Fi HaLow, is envisioned for long-range and low-power communication. It is sub-1 GHz technology designed for massive Internet of Things (IoT) and machine-to-machine devices. It aims to overcome the IoT challenges, such as providing connectivity to massive power-constrained devices distributed over a large geographical area. To accomplish this objective, IEEE 802.11ah introduces several unique physical and medium access control layer (MAC) features. In recent years, the MAC features of IEEE 802.11ah, including restricted access window, authentication (e.g., centralized and distributed) and association, relay and sectorization, target wake-up time, and traffic indication map, have been intensively investigated from various aspects to improve resource allocation and enhance the network performance in terms of device association time, throughput, delay, and energy consumption. This survey paper presents an in-depth assessment and analysis of these MAC features along with current solutions, their potentials, and key challenges, exposing how to use these novel features to meet the rigorous IoT standards.

The key attributes envisioned for LTE-Advanced pertaining to 5G Networks are ubiquitous presence, device convergence, massive machine connectivity, ultrahigh throughput and moderated carbon footprint of the network and the user equipment actuated by offloading cellular data traffic and by enabling device to device communication. The present method of mobility management and addressing as the authors have foreseen in LTE Advanced can solve some issues of cellular traffic backhaul towards the access and core network by actuating a local breakout and enabling communication directly between devices. But most of the approaches look forward towards an enhancement in the radio resource allocation process and prone to interference. Besides, most of these proposals delve in Device to Device (D2D) mode initiation from the device end, but no research has so far addressed the concept of a network initiated D2D process, which can optimise the channel utilisation and network operations further. In their attempt to knot these loose ends together, the auhtors furnish the concept of WISDOM (Wireless Innovative System for Dynamic Operating Mega communications) (Badoi Cornelia-I., Prasad N., Croitoru V., Prasad R., 2011) (Prasad R., June 2013) (Prasad R.,December 2013) and SMNAT (Sanyal, R., Cianca, E. and Prasad,R.,2012a) () () () (. Further, the authors explore how SMNAT (Smart Mobile Network Access Topology) can engage with WISDOM in cooperative communication to actuate D2D communication initiated by the device or the network. WISDOM is an architectural concept for 5G Networks based on cognitive radio approach. The cognition, sustained by adaptation techniques, is a way to provide communication, convergence, connectivity, co-operation, and content, anytime and anywhere. Though D2D communication using a dedicated spectrum in multi cell environment is possible through advanced network coding or by use of fractional frequency reuse, but physical proximity of the 2 devices is still a key requisite. In this paper the authors will discuss SMNAT which employs physical layer addressing to enable D2D communication agnostic to the spatial coordinates of the devices.

Resource allocation for machine-type communication (MTC) devices is one of the keys challenges in the 5G network as it affects the lifetime of battery powered devices and also the quality of service of the applications. MTC devices are battery restrained and cannot afford a lot of power consumption due to spectrum usage. In this paper, we propose a novel resource allocation algorithm termed threshold controlled access (TCA) protocol. We propose a novel technique of uplink resource allocation in which the devices make a decision of resource allocation blocks based on their battery status and related application’s power profile that eventually leads to required quality of service (QoS) metric. The first phase of the TCA algorithm selects the number of carriers to be allocated to a certain device for the better lifetime of low power MTC devices. In the second phase, the efficient solution is implemented through inducing a threshold value. A certain value of the threshold is selected through a mapping based on a QoS metric. The threshold enhances the selection of subcarriers for less powered devices, such as small e-health sensors. The algorithm is simulated for the physical layer of the 5G network. Simulation results show that the proposed algorithm is less complex and achieves better performance when compared to existing solutions in the literature.

Předkládaná disertační práce je zaměřena na "heterogenní propojení mobilních zařízení v bezdrátových systémech 5. generace". Navzdory nepochybnému pokroku v rámci navržených komunikačních řešení postrádají mobilní sítě nastupující generace dostatečnou šířku pásma a to hlavně kvůli neefektivnímu využívání rádiového spektra. Tato situace tedy v současné době představuje řadu otázek v oblasti výzkumu. Hlavním cílem této disertační práce je proto návrh nových komunikačních mechanismů pro komunikaci mezi zařízeními v bezprostřední blízkosti s asistencí mobilní sítě a dále pak návrh a implementace algoritmů pro dynamické přidělování frekvenčního spektra v nastupujících mobilních sítích 5G. Navrhnuté komunikační mechanismy a algoritmy jsou následně komplexně vyhodnoceny pomocí nově vyvinutých simulačních nástrojů (kalibrovaných s využitím 3GPP trénovacích dat) a zejména pak v experimentální mobilní síti LTE-A, která se nachází v prostorách Vysokého učení technického v Brně, Česká Republika. Získané praktické výsledky, které jsou podpořeny zcela novou matematickou analýzou ve speciálně navržených charakteristických scénářích, představují řešení pro vlastníka spektra v případě požadavků na jeho dynamické sdílení. Tato metoda tedy představuje možnost pro efektivnější využití spektra v rámci mobilních sítí 5G bez degradace kvality služeb (QoS) a kvality zážitků (QoE) pro koncové uživatele. Vědecký přínos dosažených výsledků dokazuje fakt, že některé z principů představených v této disertační práci byly zahrnuty do celosvětově uznávaného standardu (specifikace) 3GPP Release 12.

The next-generation electric power system, known as smart grid, relies on a robust and reliable underlying communication infrastructure to improve the efficiency of electricity distribution. Cellular networks, e.g., LTE/LTE-A systems, appear as a promising technology to facilitate the smart grid evolution. Their inherent performance characteristics and well-established ecosystem could potentially unlock unprecedented use cases, enabling real-time and autonomous distribution grid operations. However, cellular technology was not originally intended for smart grid communication, associated with highly-reliable message exchange and massive device connectivity requirements. The fundamental differences between smart grid and human-type communication challenge the classical design of cellular networks and introduce important research questions that have not been sufficiently addressed so far. Motivated by these challenges, this doctoral thesis investigates novel radio access network (RAN) design principles and performance analysis for the seamless integration of smart grid traffic in future cellular networks. Specifically, we focus on addressing the fundamental RAN problems of network scalability in massive smart grid deployments and radio resource management for smart grid and human-type traffic. The main objective of the thesis lies on the design, analysis and performance evaluation of RAN mechanisms that would render cellular networks the key enabler for emerging smart grid applications. The first part of the thesis addresses the radio access limitations in LTE-based networks for reliable and scalable smart grid communication. We first identify the congestion problem in LTE random access that arises in large-scale smart grid deployments. To overcome this, a novel random access mechanism is proposed that can efficiently support real-time distribution automation services with negligible impact on the background traffic. Motivated by the stringent reliability requirements of various smart grid operations, we then develop an analytical model of the LTE random access procedure that allows us to assess the performance of event-based monitoring traffic under various load conditions and network configurations. We further extend our analysis to include the relation between the cell size and the availability of orthogonal random access resources and we identify an additional challenge for reliable smart grid connectivity. To this end, we devise an interference- and load-aware cell planning mechanism that enhances reliability in substation automation services. Finally, we couple the problem of state estimation in wide-area monitoring systems with the reliability challenges in information acquisition. Using our developed analytical framework, we quantify the impact of imperfect communication reliability in the state estimation accuracy and we provide useful insights for the design of reliability-aware state estimators. The second part of the thesis builds on the previous one and focuses on the RAN problem of resource scheduling and sharing for smart grid and human-type traffic. We introduce a novel scheduler that achieves low latency for distribution automation traffic while resource allocation is performed in a way that keeps the degradation of cellular users at a minimum level. In addition, we investigate the benefits of Device-to-Device (D2D) transmission mode for event-based message exchange in substation automation scenarios. We design a joint mode selection and resource allocation mechanism which results in higher data rates with respect to the conventional transmission mode via the base station. An orthogonal resource partition scheme between cellular and D2D links is further proposed to prevent the underutilization of the scarce cellular spectrum. The research findings of this thesis aim to deliver novel solutions to important RAN performance issues that arise when cellular networks support smart grid communication.
Las redes celulares, p.e., los sistemas LTE/LTE-A, aparecen como una tecnología prometedora para facilitar la evolución de la próxima generación del sistema eléctrico de potencia, conocido como smart grid (SG). Sin embargo, la tecnología celular no fue pensada originalmente para las comunicaciones en la SG, asociadas con el intercambio fiable de mensajes y con requisitos de conectividad de un número masivo de dispositivos. Las diferencias fundamentales entre las comunicaciones en la SG y la comunicación de tipo humano desafían el diseño clásico de las redes celulares e introducen importantes cuestiones de investigación que hasta ahora no se han abordado suficientemente. Motivada por estos retos, esta tesis doctoral investiga los principios de diseño y analiza el rendimiento de una nueva red de acceso radio (RAN) que permita una integración perfecta del tráfico de la SG en las redes celulares futuras. Nos centramos en los problemas fundamentales de escalabilidad de la RAN en despliegues de SG masivos, y en la gestión de los recursos radio para la integración del tráfico de la SG con el tráfico de tipo humano. El objetivo principal de la tesis consiste en el diseño, el análisis y la evaluación del rendimiento de los mecanismos de las RAN que convertirán a las redes celulares en el elemento clave para las aplicaciones emergentes de las SGs. La primera parte de la tesis aborda las limitaciones del acceso radio en redes LTE para la comunicación fiable y escalable en SGs. En primer lugar, identificamos el problema de congestión en el acceso aleatorio de LTE que aparece en los despliegues de SGs a gran escala. Para superar este problema, se propone un nuevo mecanismo de acceso aleatorio que permite soportar de forma eficiente los servicios de automatización de la distribución eléctrica en tiempo real, con un impacto insignificante en el tráfico de fondo. Motivados por los estrictos requisitos de fiabilidad de las diversas operaciones en la SG, desarrollamos un modelo analítico del procedimiento de acceso aleatorio de LTE que nos permite evaluar el rendimiento del tráfico de monitorización de la red eléctrica basado en eventos bajo diversas condiciones de carga y configuraciones de red. Además, ampliamos nuestro análisis para incluir la relación entre el tamaño de celda y la disponibilidad de recursos de acceso aleatorio ortogonales, e identificamos un reto adicional para la conectividad fiable en la SG. Con este fin, diseñamos un mecanismo de planificación celular que tiene en cuenta las interferencias y la carga de la red, y que mejora la fiabilidad en los servicios de automatización de las subestaciones eléctricas. Finalmente, combinamos el problema de la estimación de estado en sistemas de monitorización de redes eléctricas de área amplia con los retos de fiabilidad en la adquisición de la información. Utilizando el modelo analítico desarrollado, cuantificamos el impacto de la baja fiabilidad en las comunicaciones sobre la precisión de la estimación de estado. La segunda parte de la tesis se centra en el problema de scheduling y compartición de recursos en la RAN para el tráfico de SG y el tráfico de tipo humano. Presentamos un nuevo scheduler que proporciona baja latencia para el tráfico de automatización de la distribución eléctrica, mientras que la asignación de recursos se realiza de un modo que mantiene la degradación de los usuarios celulares en un nivel mínimo. Además, investigamos los beneficios del modo de transmisión Device-to-Device (D2D) en el intercambio de mensajes basados en eventos en escenarios de automatización de subestaciones eléctricas. Diseñamos un mecanismo conjunto de asignación de recursos y selección de modo que da como resultado tasas de datos más elevadas con respecto al modo de transmisión convencional a través de la estación base. Finalmente, se propone un esquema de partición de recursos ortogonales entre enlaces celulares y D2

[ES] ¿Cómo es la atribución en un entorno de omnicanal? Se puede determinar una distinción importante en contraste con la atribución en un entorno multicanal. Además de proporcionar el proceso de análisis de marketing, una especificación del proceso estándar intersectorial para la minería de datos (CRISP¿DM), se utiliza un enfoque de método mixto secuencial para analizar la cuestión principal de la investigación. En el primer paso de esta investigación se analizan las características y los requisitos de atribución eficiente en un entorno omnicanal. A partir de entrevistas semiestructuradas con expertos y de un proceso de investigación bibliográfica holística estructurada, se identifica claramente la falta de un enfoque de atribución omnicanal. Los enfoques de atribución existentes se identifican mediante la realización de un proceso estructurado de revisión de la literatura. Estos enfoques identificados se evalúan aplicando los resultados de las entrevistas semiestructuradas con expertos, es decir, los requisitos y características de una atribución omnicanal eficiente. Ninguno de los enfoques de atribución identificados cumple con la mayoría de los requisitos de omnicanal analizados. Al tener la brecha de investigación ¿ la falta de un enfoque de atribución de omnicanales ¿ claramente identificada, se desarrolla un enfoque de atribución de omnicanales en la segunda parte de esta investigación presentada. Utilizando la metodología MAP, la principal laguna de investigación se llena proporcionando el Holistic Customer Journey (HCJ): una base de datos lista para el omni¿canal y un enfoque de atribución de omni¿canal correspondiente. Entre otras cosas, el enfoque de atribución desarrollado consiste en una clasificación de aprendizaje automático. Esta investigación presentada es la primera en utilizar información de casi 240.000.000 de conjuntos de datos de interacción, que contienen información entre dispositivos y entre plataformas. Todas las fuentes de datos subyacentes son proporcionadas por una de las plataformas inmobiliarias más grandes de Alemania.
[CAT] Com és l'atribució en un entorn de omnicanal? Es pot determinar una distinció important en contrast amb l'atribució en un entorn multicanal. A més de proporcionar el procés d'anàlisi de màrqueting, una especificació del procés estàndard intersectorial per a la mineria de dades (CRISP¿DM), s'utilitza un enfocament de mètode mixt seqüencial per analitzar la qüestió principal de la investigació. En el primer pas d'aquesta investigació s'analitzen les característiques i els requisits d'atribució eficient en un entorn omnicanal. A partir d'entrevistes semiestructurades amb experts i d'un procés de recerca bibliogràfica holística estructurada, s'identifica clarament la falta d'un enfocament d'atribució omnicanal. Els enfocaments d'atribució existents s'identifiquen mitjançant la realització d'un procés estructurat de revisió de la literatura. Aquests enfocaments identificats s'avaluen aplicant els resultats de les entrevistes semiestructurades amb experts, és a dir, els requisits i característiques d'una atribució omnicanal eficient. Cap dels enfocaments d'atribució identificats compleix amb la majoria dels requisits de omnicanal analitzats. En tenir la bretxa de recerca ¿ la manca d'un enfocament d'atribució de omnicanales ¿ clarament identificada, es desenvolupa un enfocament d'atribució de omnicanales a la segona part d'aquesta investigació presentada. Utilitzant la metodologia MAP, la principal llacuna de recerca s'omple proporcionant el Holistic Customer Journey (HCJ): una base de dades a punt per al omni¿canal i un enfocament d'atribució de omni¿canal corresponent. Entre altres coses, l'enfocament d'atribució desenvolupat consisteix en una classificació d'aprenentatge automàtic. Aquesta investigació presentada és la primera a utilitzar informació de gairebé 240.000.000 de conjunts de dades d'interacció, que contenen informació entre dispositius i entre plataformes. Totes les fonts de dades subjacents són proporcionades per una de les plataformes immobiliàries més grans d'Alemanya.
[EN] What does attribution in an omni¿channel environment look like? A major distinction can be determined in contrast to attribution in a multi¿channel environment. Besides providing the Marketing Analytics Process, a specification of the Cross¿industry standard process for data mining (CRISP¿DM), a sequential mixed method approach is utilized to analyze the main research question. Within the first step of this presented research characteristics, and requirements of efficient attribution in an omni¿channel environment are analyzed. Based on semi¿structured expert interviews and a holistic structured literature research process, the lack of an omni¿channel attribution approach is clearly identified. Existing attribution approaches are identified by conducting the structured literature review process. Those identified approaches are evaluated by applying the results of the semi¿structured expert interviews - the requirements and characteristics of efficient omni¿channel attribution. None of the identified attribution approaches fulfill a majority of the analyzed omni¿channel requirements. By having the research gap - the lack of an omni¿channel attribution approach - clearly identifed, an omni¿channel attribution approach is developed in the second part of this presented research. Utilizing the MAP methodology, the main research gap is filled by providing the Holistic Customer Journey (HCJ): an omni¿channel ready data foundation and a corresponding omni¿channel attribution approach. Among other things the developed attribution approach consists of a machine learning classification. This presented research is the first to utilize information from almost 240.000.000 interaction data sets, containing crossdevice and cross¿platform information. All underlying data sources are provided by one of Germany's largest real¿estate platforms.
Nass, O. (2019). Optimizing Online Marketing Efficiency By Analyzing the Mutual Influence of Online Marketing Channels with Respect to Different Devices [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/122296
TESIS

Machine-to-Machine (M2M) communications represent the cornerstone technology enabling the pervasive deployment of automated applications connecting billions of devices or objects without the need for human intervention. Within the past few years, the number of M2M-based services has dramatically grown, spurred by remarkable benefits provided by the development of smart and cost-effective applications in a wide range of areas, including remote sensing, health monitoring and Intelligent Transportation Systems (ITS). Therefore, great research effort has been invested to the design of novel solutions or to the optimization of existing communication systems in order to accommodate M2M traffic with diverse QoS requirements. This thesis introduces novel approaches to enhance M2M communications based on the adoption of Device-to-Device (D2D) connectivity among the devices, in conjunction with existing techniques widely implemented, such as clustering, packet aggregation and trunking. D2D communications are an attractive and intelligent solution to reduce the power consumption and the latency due to the short-range connectivity employed by two devices operating within the licensed spectrum. In the first approach, we envision a centralized scheme to allocate cellular resources in a scenario where the machine devices can exploit D2D connectivity, realizing a multi-hop D2D network overlaying an LTE-A system. The main challenges involving the D2D link establishment are covered, such as proximity discovery and device pairing, and a routing mechanism is proposed in order to efficiently forward packets to a collector device through multi-hop transmissions. To address the RAN overload issue originated from the mass access, machine devices can communicate with a cellular user in the proximity through low power D2D connections a cellular user can collect and combine the traffic generated by machine devices through low power D2D connections and then forward the aggregated data to the attached base station, along with the traffic originated from higher layers. As a result, enormous benefits can be obtained in terms of energy efficiency and throughput, as the cellular acts as a relay and is responsible of the data transmission on the trunked cellular uplink. We further enhance this mechanism by introducing a backoff mechanism to mitigate collisions occurring during the access reservation phase, performed by the machine devices before transmitting a packet. In addition, we evaluate the advantages in terms of trunking gains provided by a multi-cluster configuration, where multiple cellular users forward traffic sent by devices belonging to different clusters, without generating interference. In the context of vehicular networking, M2M communications aim to enable the internet-working among connected vehicles, which can autonomously exchange information and make critical decisions in diverse situations. The main challenge associated with Vehicular Ad-hoc Networks (VANET) is how to provide reliable connectivity between vehicles in scenarios with highly dynamic topologies and unpredictable channel conditions. We choose to overcome these issues by developing a clustering algorithm based on the mobility correlation degree of the vehicles in a typical highway scenario, along with a relaying scheme aiming at supporting the communication between different clusters. The proposed protocol is then evaluated by implementing a realistic mobility model, taking into account interactions between approaching vehicles and assuming that the communications are established according to the IEEE 802.11p/WAVE standard.

Device-to-device (D2D) communication has emerged as a promising concept for supporting the vehicular networks, which can efficient and reliable enhance cellular network. In this paper, we discuss two different design criteria for vehicular networks. They are maximizing overall vehicle-to-infrastructure (V2I) link throughput while guaranteeing the minimum reliability for each vehicle-to-vehicle (V2V) link, and maximizing the minimum throughput of all V2V links under the constraints of minimum V2I link throughput requirements. Because both of these problems is an mixed integer non-linear programming problem, we solve these problems in two steps, i.e., by first clustering D2D users into clusters and then optimizing their respective power allocations. Specifically, we first propose a spectral clustering (SC) method for D2D users clustering. Then, two power allocation algorithms are developed to maximize the sum V2I link throughput and maximize the minimum V2V link throughput, respectively. The effectiveness of proposed resource allocation algorithms is validated by computer simulation.

Researchers are attracted to emerging field 5G with machine learning. Many review articles have been carried out to analyze in a different direction of 5G with machine learning. However, no researcher presented bibliometric analysis on machine learning in the 5G research field to a detailed analysis of research status and future trend network in this research area. A bibliometric analysis was done in the current study using the bibliometric R tool and VOS viewer software. The relevant literature was collected period 2001 to 2021 from the Web of Science (WoS) Core Collection and Scopus database. The quantitative analysis was done in terms of a yearly published article, most trend research topic, and future direction in ML in 5G technology. Finally, the result indicated that China, the U.S.A., and India are the top countries to publish this field because China, the U.S.A., and the U.K. are the most cited countries. Beijing University of Posts and Telecommunications is the most relevant organization, Wang most appropriate and most influential author in this research area (5G in AI/ML). IEEE Access, IEEE transactions on vehicular technology, and Sensor are the most relevant journal. The main challenges in this field are low latency communication, resource allocation, resource management spectral efficiency, millimeter wave, 5G with the Internet of things (IoT), a device to device communication, power control, and massive MIMO. Deep learning, machine learning, cognitive radio, and reinforcement learning are artificial intelligence techniques used in 5G.

Converging Cloud computing with Internet of Things transformed organizations' traditional technologies. This chapter examines the intersection of cloud computing and internet of things in consort with how these solutions often interact on the internet. Vendors develop CloudIoT capabilities to support organizations' day-to-day operations. IoT is a combined platform encompassing physical and virtual nodes. IoT objects comprise device-to-device data sharing, machine-to-machine provisioning, sensors, actuators, and processors. These systems may be deployed as hardware components and applications software. This chapter also emphasizes data security, reliability, resource provisioning, service-level agreement, quality of service, IoT, privacy, and device integration. This chapter also highlights operational benefits and/or security issues affecting CC and IoT technologies.

With the growth of Internet of Things and user demand for personalized applications, context-aware applications are gaining popularity in current IT cyberspace. Personalized content, which can be a notification, recommendation, etc., are generated based on the contextual information such as location, temperature, and nearby objects. Furthermore, contextual information can also play an important role in security management of user or device in real time. When the context of a user or device changes, the security mechanisms should also be updated in real time for better performance and quality of service. Access to a specific resource may also be dependent upon user's/device's current context. In this chapter, the role of contextual information for IoT application security is discussed and a framework is provided which auto-updates security policy of the device based on its current context. Proposed framework makes use of machine learning algorithm to update the security policies based on the current context of the IoT device(s).

With a greater boost in the development of technology, the healthcare industry has had a huge development in terms of offering innovative diagnoses and services to patients. Healthcare has become an indispensable resource in the daily lives of humans leading to a huge innovation in the healthcare sector. To improve the efficiency of diagnosis and with the potential to offer better treatment to patients, the healthcare industry has adopted numerous technologies such as the internet of things (IoT), machine learning, blockchain, etc., to pave way for sustainability in the medicinal sector. These systems, though very advanced, have some drawbacks; like problems of security and privacy of medical data that can expose these records to threats of privacy, and loss of data through manhandling of records and data transactions resulting in delays, while the patient is being monitored. etc., which can pose a major threat to the patient's well-being. This paper will provide how a blockchain network in any IoT-based device can help in the preservation of these medical records, using suitable algorithms.

It is suggested by many scholars that if the goal of engineering education is to produce engineers who can critically design and create, then providing students with early opportunities to engage in creative engineering design is important. While basic design is focused on the development of new products for the individual, working towards a more sustainable world demands greater attention to designing for and with communities. Improving design education and examining design-learning outcomes requires a kind of targeted approach that could match the best practices to personalize student learning. Design is complex and design includes balancing the needs of multiple stakeholders. However, there is a gap in the preparation of design education that will be needed in a challenging environment. This paper reviews the history of design thinking in the engineering curriculum. Design thinking education starts with an understanding of its importance with socioeconomic relevance. Through observation and empathy, mapping the designer uses the listening and learning tools for mapping users unarticulated needs, working in a team environment. The designer takes time to think carefully why a certain project is considered and details which aspects of machine learning application can be applied from functional to complete success for the end users. The availability of powerful virtual reality methodologies, have made it possible to consider the realistic needs and visualize scenarios and to explore the design alternatives with new ideas before full scale resource allocation on new ideas. Mid-to-advanced level courses with experimental assignments require that students apply through experimentation the principles and concepts learned in foundation courses. The basic design tools such as axiomatic thinking, theory of inventive problem solving, design iteration and simulation using hardware-in-the loop are discussed with case studies. Consideration of product sustainability with the thoughts of design for disassembly and disposal has emerged as a major part of design thinking. Senior engineering courses center on cross and interdisciplinary design and capstone experiences so that students experience fully guided practice of device design and problem solving, simulating what they are likely to experience in the world. This paper examines the critical issues of design thinking in a curriculum from observation, empathy mapping, validation of the idea, and improvement of idea by virtual reality and machine learning, optimization of the idea by tools such as axiomatic design, hardware in the loop simulation, and finally examining product sustainability causes.

The network is very important to the normal operation of all aspects of society and economy, and the memory leak of network device is a software failure that seriously damages the stability of the system. Some common memory checking tools are not suitable for network devices that are running online, so the operation staff can only constantly monitor the memory usage and infer from experience, which has been proved to be inefficient and unreliable. In this paper we proposed a novel memory leak detection method for network devices based on Machine learning. It first eliminates the impact of large-scale resource table entries on the memory utilization. Then, by analyzing its monotonicity and computing the correlation coefficient with the memory leak sequence sets pre constructed by simulation, the memory leak fault can be found in time. The simulation experiments show that the scheme is computationally efficient and the precision rate is close to 100%, it works well in the actual network environment, and has excellent performance.

Abstract Despite the outstanding improvements achieved by artificial intelligence in the Structural Health Monitoring (SHM) field, some challenges need to be coped with. Among them, the necessity to reduce the complexity of the models and the data-to-user latency time which are still affecting state-of-the-art solutions. This is due to the continuous forwarding of a huge amount of data to centralized servers, where the inference process is usually executed in a bulky manner. Conversely, the emerging field of Tiny Machine Learning (TinyML), promoted by the recent advancements by the electronic and information engineering community, made sensor-near data inference a tangible, low-cost and computationally efficient alternative. In line with this observation, this work explored the embodiment of the One Class Classifier Neural Network, i.e., a neural network architecture solving binary classification problems for vibration-based SHM scenarios, into a resource-constrained device. To this end, OCCNN has been ported on the Arduino Nano 33 BLE Sense platform and validated with experimental data from the Z24 bridge use case, reaching an average accuracy and precision of 95% and 94%, respectively.

Reliability and Maintainability analyses are becoming an increasing competitive advantage in machine tool design. In particular, the goal of machine tools for Ultra High Precision Machining is to guarantee high specified performances and to maintain them over life cycle time. A structured reliability approach applied to such complex and innovative systems must be integrated in the early phase of the design. In this paper, the reliability characterization of an adjustable platform for micromilling operations is presented. The platform is intended to improve the surface finishing of the workpiece, through a broadband Active Vibration Control device based on high performance piezoelectric multilayer actuators. The study intends to assess the capability of the system to maintain along the life cycle the appropriate reduction of the chattering vibrations without any shape error. By dividing the system through a morphological-functional decomposition, the critical elements are detected and their reliability issues are extensively discussed. Their lifetimes are described through opportune distributions and models. The study is completed by the quantitative reliability prediction of the overall system. Finally, a sensitivity analysis is performed and reliability allocation implications are evaluated to determine the effect of every component on the system reliability characteristics and life cycle cost.

In this work it is proposed to use special lubricant materials with the properties manageable by electric field. Such lubricant materials allow to provide the viscosity needed for liquid friction adaptive to the conditions of the loading interaction of the contacting interfaces of bearings and sliding supports. The coefficients of friction were evaluated by means of modified friction machine with planar parallel scheme of the sample interaction realization. At that the sliding speed of the sample relative to the counter body increased smoothly according to the sinusoidal law from zero up to maximum value. Then it decreased smoothly to the full stop and the movement reversed; contact pressures after working out the given number of cycles increased stepwise from 0.5 to 3.2 MPa. The range of the optimal intensities lain in between 120–150V. An increase in the relative resource of efficiency of the frictional device up to 3 times was found in these conditions.

Special considerations must be given in medical air transportation to medical professionals, medical equipment, and flight crews due to the conditions of the critically ill patients transported, the intra-aircraft environment, and the need to make emergency flights. However, few studies have investigated the ergonomic problems faced by the medical air transportation services provided in Japan. Therefore, here we investigated ergonomic problems, including those related to medical device use, faced by medical personnel and the aviation community in Japan. Results indicated that basic education and training on aerospace physiology, intra-aircraft use of medical equipment, flight operation systems, and crew resource management are currently insufficient in standard medical education. In fact, most medical devices used during flights are conventional devices with no ergonomic considerations made for vibration or low cabin pressure. In particular, problems relating to human-machine interfaces, power supplies, electromagnetic compatibility, and the ergonomic and technical compatibilities of electronic medical equipment used during the flights must be improved. Improvements are also needed in relation to safety during emergency flights such as those made at night or in bad weather, crew training, and air traffic control systems during large-scale disasters. Interdisciplinary collaboration is required to further investigate and resolve these problems.