This thesis has been carried out on behalf of the department of Digital Services - SDI at Ericsson. Ericsson Software Defined Infrastructure (SDI) is a telco grade hardware management solution for cloud infrastructure. In datacenter deployments, the extra management equipment needed by the solution becomes insignificant due to the amount of equipment it manages. But the closer to the cloudedge you get, the smaller in size the deployments become, thus making the management equipment an ever-increasing share of the total deployment seize leading to inefficient resource utilization, so called overhead. Especially with Distributed Radio Access Networks (RAN) many small deployments, often only consisting of a single compute-server used to process radio, will be deployed at radio sites and/or in buildings around cities to deliver cell service. In this type of usage, the overhead of equipment management builds up cumulatively due to the numerous amounts of deployments. This overhead leads to excessive maintenance, power usage, space needs for equipment, costs, and electronic waste. The goal of this thesis was to evaluate how to reduce the overhead of equipment management in a scenario involving numerous small-capacity widely-distributed sites which are common in the 5G telco cloud. The idea was to determine if the overhead could be reduced by exploiting Baseboard Management Controllers (BMC), this was tested by designing a low-footprint and lightweight proof of concept equipment management solution and implementing a prototype of it. By testing, verifying, and analyzing the proof-of-concept solution, it was concluded that by exploiting the BMC to run a custom software service that phoned home to a centralized management server it was possible to drastically reduce the overhead in such scenarios. It also became clear that BMCs could have even more usage areas and provide even greater value if support to run third partyapplications existed among them. / Detta examensarbete har utförts på uppdrag av avdelningen Digital Services - SDI på Ericsson. Ericsson Software Defined Infrastructure (SDI) är en hårdvaruhanteringslösning av telekomkvalitetför molnifrastruktur. I datacenterinstallationer blir den extra hanteringsutrustning som behövs av lösningen obetydlig på grund av mängden utrustning den hanterar. Men ju längre ut till molnkanten du kommer, desto mindre blir storleken på installationerna, vilket gör att hanteringsutrustning blir en ständigt ökande andel av den totala installationsstorleken som leder till ineffektivt resursutnyttjande, så kallat overhead. Speciellt med Distribuerade Radio Access Nätverk (RAN) så kommer många små installationer, ofta endast bestående av en enda server som utför radiolänksberäkningar, att vara fysiskt utplacerade vidbasstationer och/eller i byggnader runt städer för att leverera och tillhandage mobiltjänster. Vid denna typ av användning ökar overheadet för utrustningshanteringen kumulativt på grund av antalet installationer. Detta overhead leder till mer underhåll och elektroniskt avfall, större utrymmesbehov för utrustning samt högre strömförbrukning och kostnader. Målet med detta examensarbete var att utvärdera hur man kan minska overheadet hosutrustningshantering när det tillämpas på ett stort antal, små decentraliserade distribuerade installationer, vilket är förekommande i telekommoln. Idén var att undersöka om overheadet kunde minskas genom att utnyttja Baseboard Management Controllers (BMC), detta testades genom att designa en liten och lättviktigt konceptlösning för utrustningshantering samt implementera en prototyp av den. Genom att testa, verifiera och analysera konceptlösningen drogs slutsatsen att det var möjligt att drastiskt minska overheaden i sådana scenarion genom att utnyttja BMC att köra en egen mjukvarutjänst, som automatiskt anslöt till en central hanteringsserver. Genom arbetet blev det också tydligt att BMC:er skulle kunna ha ännu fler användningsområden och ge ännu större värde om stödet för att köra tredjepartsapplikationer på dem var mer utbrett

The maturing capabilities of Artificial Intelligence (AI) and Machine Learning (ML) have resulted in increased attention in research and development on adopting AI and ML in 5G and future networks. With the increased maturity, the usage of AI/ML models in production is becoming more widespread, and maintaining these systems is more complex and likely to incur technical debt when compared to standard software. This is due to inheriting all the complexities of traditional software in addition to ML-specific ones. To handle these complexities the field of ML Operations (MLOps) has emerged. The goal of MLOps is to extend DevOps to AI/ML and therefore speed up development and ease maintenance of AI/ML-based software, by, for example, supporting automatic deployment, monitoring, and continuous re-training of models. This thesis investigates how to construct an MLOps workflow by selecting a number of tools and using these to implement a workflow. Additionally, different approaches for triggering re-training are implemented and evaluated, resulting in a comparison of the triggers with regards to execution time, memory and CPU consumption, and the average performance of the Machine learning model.

By 2020 there will be 50 billion connected devices over the Internet. With the fast-increasing data traffic demand in both fixed and mobile networks, network operators need to migrate networks towards next generation solutions. The network migration requires the enormous investment in equipment and infrastructure, while the revenues are not expected to grow significantly. Therefore, one of the main challenges for network operators is to find out a proper cost-effective optical network solution that can match future high capacity demand and flexibly support multiple network services on a common network infrastructure. The first part of the thesis addresses the Active Optical Network (AON) and its migration strategies towards Next Generation Optical Access (NGOA) solutions. Several migration strategies are proposed from the perspective of network topology, data plane and control plane. A general methodology for Techno-Economic analysis has been developed and applied to the Total Cost of Ownership (TCO) calculation of different NGOA solutions. The thesis provides a complete cost evaluation of AON migration paths, which can be used by network operators to assess the economic feasibility of network migration. A converged Optical Transport Network (OTN) that can serve both fixed and mobile network services is beneficial from the cost-saving perspective. However, the different types of services, require different network performance. The second part of the thesis focuses on the investigation of the converged OTN that can be flexibly and timely adjusted to satisfy varying service conditions. A programmable OTN featured with Wavelength Division Multiplexing (WDM) in the data plane and Software Defined Networking (SDN) in control plane has been proposed. To demonstrate the benefits of the converged OTN, the thesis also provides a multi-domain orchestration architecture for the multiple network services.  The resource orchestration, across three network domains: OTN, mobile network and cloud, enables agile service creation and optimized resource allocation among the multiple domains. / <p>QC 20170512</p>

O presente trabalho descreve o desenvolvimento e implementação de duas configurações distintas de um receptor optoeletrônico integrado. A primeira configuração é similar a um projeto encontrado na literatura mas apresenta diversas modificações que lhe conferiram melhor desempenho em comparação ao projeto original. A segunda configuração é uma nova proposta deste trabalho. O receptor foi desenvolvido e implementado visando sua aplicação em redes de comunicações ópticas passivas (PONs) de alta velocidade comutadas a pacote, para possibilitar a utilização da técnica de multiplexação em comprimento de onda (WDM), aumentando assim a capacidade de transmissão da rede, em particular no ramo de ligação da rede de serviços com o usuário final, denominado rede de acesso. O principal objetivo do receptor aqui desenvolvido foi proporcionar uma sintonia rápida entre os canais disponíveis na rede, possibilitando sua seleção num tempo inferior àquele necessário para a transmissão de um único pacote de informação, diminuindo assim o atraso de sintonia e, por conseguinte, a perda de informação. Para tanto, os circuitos integrados implementados e caracterizados referem-se aos circuitos de chaveamento eletrônico e do amplificador de transimpedância das duas configurações investigadas. Os dados experimentais obtidos para as duas configurações confirmaram a previsão de chaveamento dos canais de entrada num intervalo de tempo da ordem de alguns nanosegundos, o que é totalmente compatível com a velocidade de transmissão das aplicações a que se destina este receptor (aproximadamente 5 Gbits/s). Adicionalmente, são apresentados os dados experimentais relativos à freqüência de corte, ganho direto, isolação, relação on/off e características de ruído dos circuitos implementados. / The present work describes the design and implementation of two configurations of an integrated optoelectronic receiver. The first one is similar to a previously reported design but with some modifications to improve its performance. The second one is a new proposal of this work. The goal of the receiver design and implementation was its application in high bit rate packet-switched passive optical networks (PONs) employing the wavelength division multiplexing (WDM) technique to increase the network capacity, in particular on the connection branch of the network core with the final user, the access network. The main goal of the receiver design was to achieve a fast channel tuning, allowing a tuning time smaller than the required for the transmission of a single information packet, decreasing the tuning latency and, therefore, the rate of information packet loss. In order to accomplish this goal, the implemented and tested integrated circuits include the electronic switching circuit and the transimpedance amplifier for both configurations investigated. The measured data for both configurations confirm the expected input channel switching time results, of about a few nanoseconds, which is certainly useful for the expected bit rate of operation (approximate 5 Gbps). Additionally, experimental results concerning cutoff frequency and bandwidth, direct gain, isolation, on/off ratio, and noise characteristics of both implemented circuits are presented.

O presente trabalho descreve o desenvolvimento e implementação de duas configurações distintas de um receptor optoeletrônico integrado. A primeira configuração é similar a um projeto encontrado na literatura mas apresenta diversas modificações que lhe conferiram melhor desempenho em comparação ao projeto original. A segunda configuração é uma nova proposta deste trabalho. O receptor foi desenvolvido e implementado visando sua aplicação em redes de comunicações ópticas passivas (PONs) de alta velocidade comutadas a pacote, para possibilitar a utilização da técnica de multiplexação em comprimento de onda (WDM), aumentando assim a capacidade de transmissão da rede, em particular no ramo de ligação da rede de serviços com o usuário final, denominado rede de acesso. O principal objetivo do receptor aqui desenvolvido foi proporcionar uma sintonia rápida entre os canais disponíveis na rede, possibilitando sua seleção num tempo inferior àquele necessário para a transmissão de um único pacote de informação, diminuindo assim o atraso de sintonia e, por conseguinte, a perda de informação. Para tanto, os circuitos integrados implementados e caracterizados referem-se aos circuitos de chaveamento eletrônico e do amplificador de transimpedância das duas configurações investigadas. Os dados experimentais obtidos para as duas configurações confirmaram a previsão de chaveamento dos canais de entrada num intervalo de tempo da ordem de alguns nanosegundos, o que é totalmente compatível com a velocidade de transmissão das aplicações a que se destina este receptor (aproximadamente 5 Gbits/s). Adicionalmente, são apresentados os dados experimentais relativos à freqüência de corte, ganho direto, isolação, relação on/off e características de ruído dos circuitos implementados. / The present work describes the design and implementation of two configurations of an integrated optoelectronic receiver. The first one is similar to a previously reported design but with some modifications to improve its performance. The second one is a new proposal of this work. The goal of the receiver design and implementation was its application in high bit rate packet-switched passive optical networks (PONs) employing the wavelength division multiplexing (WDM) technique to increase the network capacity, in particular on the connection branch of the network core with the final user, the access network. The main goal of the receiver design was to achieve a fast channel tuning, allowing a tuning time smaller than the required for the transmission of a single information packet, decreasing the tuning latency and, therefore, the rate of information packet loss. In order to accomplish this goal, the implemented and tested integrated circuits include the electronic switching circuit and the transimpedance amplifier for both configurations investigated. The measured data for both configurations confirm the expected input channel switching time results, of about a few nanoseconds, which is certainly useful for the expected bit rate of operation (approximate 5 Gbps). Additionally, experimental results concerning cutoff frequency and bandwidth, direct gain, isolation, on/off ratio, and noise characteristics of both implemented circuits are presented.

Le streaming adaptatif sur HTTP, désigné par HAS, est une technique de streaming vidéo largement déployée sur Internet. Elle utilise TCP comme protocole de transport. Elle consiste à segmenter la vidéo stockée sur un serveur web en petits segments indépendants de même durée de lecture et transcodés à plusieurs niveaux de qualité, désignés par "chunks". Le player, du côté du client HAS, demande périodiquement un nouveau chunk. Il sélectionne le niveau de qualité en se basant sur l’estimation de la bande passante du/des chunk(s) précédent(s). Étant donné que chaque client HAS est situé au sein d’un réseau d’accès, notre étude traite un cas fréquent dans l’usage quotidien: lorsque plusieurs clients partagent le même lien présentant un goulot d’étrangement et se trouvent en compétition sur la bande passante. Dans ce cas, on signale une dégradation de la qualité d’expérience (QoE) des utilisateurs de HAS et de la qualité de service (QoS) du réseau d’accès. Ainsi, l’objectif de cette thèse est d’optimiser le protocole TCP pour résoudre ces dégradations de QoE et QoS. Notre première contribution propose une méthode de bridage du débit HAS au niveau de la passerelle. Cette méthode est désignée par "Receive Window Tuning Method" (RWTM): elle utilise le principe de contrôle de flux de TCP et l’estimation passive du RTT au niveau de la passerelle. Nous avons comparé les performances de RWTM avec une méthode récente implémentée à la passerelle qui utilise une discipline particulière de gestion de la file d’attente, qui est désignée par "Hierarchical Token Bucket shaping Method" (HTBM). Les résultats d’évaluations indiquent que RWTM offre une meilleure QoE et une meilleure QoS de réseau d’accès que HTBM. Notre deuxième contribution consiste à mener une étude comparative combinant deux méthodes de bridages, RWTM et HTBM, avec quatre variantes TCP largement déployées, NewReno, Vegas, Illinois et Cubic. Les résultats d'évaluations montrent une discordance importante entre les performances des différentes combinaisons. De plus, la combinaison qui améliore les performances dans la majorité des scénarios étudiés est celle de RWTM avec Illinois. En outre, une mise à jour efficace de la valeur du paramètre "Slow Start Threshold", sthresh, peut accélérer la vitesse de convergence du player vers la qualité optimale. Notre troisième contribution propose une nouvelle variante de TCP adaptée aux flux HAS, qu’on désigne par TcpHas; c’est un algorithme de contrôle de congestion de TCP adapté aux spécifications de HAS. TcpHas estime le niveau de la qualité optimale du flux HAS en se basant sur l’estimation de la bande passante de bout en bout. Ensuite, TcpHas applique un bridage au trafic HAS en fonction du débit d’encodage du niveau de qualité estimé. TcpHas met à jour ssthresh pour accélérer la vitesse de convergence. Une étude comparative a été réalisée avec la variante Westwood+. Les résultats d’évaluations montrent que TcpHas est plus performant que Westwood+. / HTTP adaptive streaming (HAS) is a streaming video technique widely used over the Internet. It employs Transmission Control Protocol (TCP) as transport protocol and it splits the original video inside the server into segments of same duration, called "chunks", that are transcoded into multiple quality levels. The HAS player, on the client side, requests for one chunk each chunk duration and it commonly selects the quality level based on the estimated bandwidth of the previous chunk(s). Given that the HAS clients are located inside access networks, our investigation involves several HAS clients sharing the same bottleneck link and competing for bandwidth. Here, a degradation of both Quality of Experience (QoE) of HAS users and Quality of Service (QoS) of the access network are often recorded. The objective of this thesis is to optimize the TCP protocol in order to solve both QoE and QoS degradations. Our first contribution consists of proposing a gateway-based shaping method, that we called Receive Window Tuning Method (RWTM); it employs the TCP flow control and passive round trip time estimation on the gateway side. We compared the performances of RWTM with another gateway-based shaping method that is based on queuing discipline, called Hierarchical Token Bucket shaping Method (HTBM). The results of evaluation indicate that RWTM outperforms HTBM not only in terms of QoE of HAS but also in terms of QoS of access network by reducing the queuing delay and significantly reducing packet drop rate at the bottleneck.Our second contribution consists of a comparative evaluation when combining two shaping methods, RWTM and HTBM, and four very common TCP variants, NewReno, Vegas, Illinois and Cubic. The results show that there is a significant discordance in performance between combinations. Furthermore, the best combination that improves performances in the majority of scenarios is when combining Illinois variant with RWTM. In addition, the results reveal the importance of an efficient updating of the slow start threshold value, ssthresh, to accelerate the convergence toward the best feasible quality level. Our third contribution consists of proposing a novel HAS-based TCP variant, that we called TcpHas; it is a TCP congestion control algorithm that takes into consideration the specifications of HAS flow. Besides, it estimates the optimal quality level of its corresponding HAS flow based on end-to-end bandwidth estimation. Then, it permanently performs HAS traffic shaping based on the encoding rate of the estimated level. It also updates ssthresh to accelerate convergence speed. A comparative performance evaluation of TcpHas with a recent and well-known TCP variant that employs adaptive decrease mechanism, called Westwood+, was performed. Results indicated that TcpHas largely outperforms Westwood+; it offers better quality level stability on the optimal quality level, it dramatically reduces the packet drop rate and it generates lower queuing delay.

The thesis sums up the traffic parameters for the evaluation of next-generation access networks. There are mentioned parameters for service quality management and parameters for measuring the quality of experience. The EU regulation on High Speed Internet in the Czech Republic is analyzed. The practical part compares the results between the implementation of the quality management of services and without implementation. Also there is comparing theoretical and practical transmission rates of physical ports. Part of the thesis is a questionnaire focused on the quality of experience of the video sequence.

The main goal of this diploma thesis was to implement the proposal of optical access networks in the selected location. Introduce is focused on the physical layer of network. This includes optical fibers, optical cables, connectors, optical splitters, and more. According to the place of termination of optical fiber will be studied best FTTx connections. According to network topology distinguished point to point and point to multipoint network. Networks points to point are also known as an active optical network and for each user are restricted to a separate fiber. Point to multipoint networks known as passive optical network. In terms of the PON is defined by many standards. For a selected location will be choice the most appropriate solution. Finally, the measurement will be made and will be considered for monitoring the route. Before the design of optical access networks, I made an illustrative simulation in program OptSim. Performed simulations helps to study the transmission characters of optical networks, depending on the physical architecture. In general, for the design of high-capabilities networks for large companies will be more appropriate active topology that provides due wavelength multiplexes high transmission capacity. To join in the urban areas with a large number of separate houses, it is more appropriate use of passive topology. For the design of optical networks has been selected location with multiple housing buildings. The end customer is not given the requirement for high data transfer rates as large companies. For the proposal was selected FTTB networks which is suitable because of lower costs. The second variant of the proposal was a passive optical network according to standard GPON. This standard was chosen because of the high splitting ratio. For both solutions was developed the entire proposal by ITU-T with a detailed drawing documentation. In the selected area was conducted measuring the optical networks in order to verify the possibility of extension actual solutions. For a sample issue of measurements, I made several measurements using optical reflectometer over an optical splitter in order to prove the need for measurements during construction. FTTH networks gaining in popularity more and more, but price is the limiting factor.

Dans ce travail de thèse, un aperçu général sur les systèmes PON est présenté et sont étudiés les mécanismes et classification PON existants. Après, nous introduisons notre première contribution qui est un algorithme d'allocation dynamique de bande passante pour EPON. Cet algorithme proposé est désigné comme «hcDBA». Par la suite, une amélioration de l'algorithme de hcDBA avec mécanisme de prédiction précoce est présentée. Notre simulation montre bien que notre algorithme hcDBA est performant comparé aux mécanismes existants en termes de taux de perte de paquets et de délai d’accès. Dans notre seconde contribution, nous sommes intéressés au problème de consommation d’énergie qui est un sujet d’actualité dans les réseaux de télécommunication. Les études montrent aujourd’hui que les réseaux d'accès constituent une partie remarquable de la consommation totale d'énergie dans les réseaux de télécommunication. Les organisations ITU-T et IEEE ont publié la recommandation pour la conversation de l'énergie pour les réseaux PONs. Bien que, la consommation totale d'énergie des nœuds ONU est plus importantes que d'autres équipements dans le réseau d'accès fixe, les normes et la plupart des travaux de recherches ont porté sur les économies d'énergie du côté de ONU. Dans cette thèse, nous sommes concentrés sur une méthode d'efficacité énergétique basée sur la conservation de l'énergie du côté de l’OLT. La méthode proposée permet d’économiser de l'énergie en déplaçant dynamiquement des cartes d’OLT en mode de sommeil profond en fonction des charges de trafic entrant et sortant / In this thesis, general overview about PON systems is presented and existing PON mechanisms and classification are investigated. After, a novel dynamic bandwidth allocation algorithm for EPON is introduced. This proposed algorithm is named as “Half Cycling Dynamic Bandwidth Allocation-hcDBA” by the inspiration of its half cycling processing mode. Later, an improvement of hcDBA algorithm with early prediction mechanism is presented. As a result statement of the study, hcDBA algorithm performs better than existing mechanism in terms of packet loss ratio and access delays. Beside, simulation traffic behavior of EPON’s upstream channel has been investigated in order to support the decision of selecting suitable traffic generator in further studies. Energy conversation is one of the hot topics in telecommunication networks. Access networks constitute remarkable portion of the total energy consumption in telecommunication networks. ITU-T and IEEE organizations published recommendation for energy conversation in PONs. While, total energy consumption of ONUs is more than other equipment in fix access network the standards and most of the researches focused on saving energy at ONU side. In this thesis I focused on an energy efficiency method based on energy conversation on OLT side. The proposed method save energy by dynamically moving OLT cards to deep sleep mode according to the incoming and outgoing traffic loads

[pt] Para suportar os requisitos das redes móveis de 5a geração (5G), os sistemas de comunicação óptica serão usados nas redes de acesso. Isso ocorre porque a evolução das RAN (Radio Access Networks) incluem a centralização do equipamento mais crítico para implantar pontos de acesso móveis de baixa potência, como DAS (Distributed Antenna Systems) e Small Cells. Os serviços emergentes solicitam a implantação de tecnologias de rádio sobre fibra com ênfase na eficiência de largura de banda, eficiência energética e alta confiabilidade. Neste âmbito, um monitoramento eficiente da camada física é imperativo para a operação dessas redes. O sistema de monitoramento deve fornecer uma localização de falhas em serviço, econômico, centralizado e com impacto mínimo para a transmissão de dados. Esta tese propõe vários sistemas de monitoramento incorporado no transceptor utilizando reflectometria óptica no domínio do tempo. Os sistemas de monitoramento são testados em diferentes sistemas de transmissão de dados e arquiteturas de rede, onde é apresentada uma validação simulada e outras experimentais. / [en] In order to support the requirements for 5th generation mobile networks (5G), optical communication systems will be used in the access part of the network. This is because the evolution of radio access networks includes the centralization of the most critical equipment in order to deploy low power mobile access points, like distributed antenna systems and small cells. The emerging services call for the deployment of radio over fibre technologies with emphasis on bandwidth efficiency, energy efficiency and high reliability. Within this scope, an efficient monitoring of the physical layer would become essential for the operation of these networks. The monitoring system should provide in-service, cost efficient and centralized fault localization with minimum impact on data transmission. This thesis proposes several transceiver-embedded optical time domain reflectometry monitoring systems. The monitoring systems are tested over different data transmission systems and network architectures, where one architecture was simulated and several others experimentally validated.