Mennan Selimi

Short‐term temperature forecasts are useful in many IoT systems, but the models behind them are usually tested on one large public dataset rather than on the cheap sensors actually deployed. We ran one forecasting pipeline—a naive “no change” baseline and three machine‐learning models (Ridge regression, Random Forest, and XGBoost) over lagged and rolling features—on two sources: the Jena Climate benchmark (Germany, 2009–2016, about 420 000 records) and 1 month of data from each of three low‐cost optical particle‐counter nodes in Tetovo, North Macedonia. All splits were chronological, and every evaluation was repeated with walk‐forward cross‐validation. Two of the three field nodes came close to the benchmark in absolute error (RMSE 0.76°C and 0.78°C, against 0.63°C on Jena) on roughly one hundredth of the training data, and two exported two‐feature models ran in real time on an Arduino MKR1000. The rest of the results were less comfortable. At one of those two nodes, no trained model beat the naive baseline, and walk‐forward evaluation showed the tree ensembles there to be far more unstable than the single split suggested. The third node's much larger error was not a modeling failure: its housing was heating in the sun, which we traced in the raw files and the site operators confirmed. Two further data‐quality defects, one per dataset, each changed a reported error figure. On this evidence, a benchmark score on its own says rather little about how a forecasting pipeline will behave on cheap edge hardware in the field.

Community networks (CNs) have gained momentum in the last few years in response to the growing demand for network connectivity in rural and urban areas. These networks, owned and managed by volunteers, offer various services to their members. While Internet access is the most popular service offered to their members, the provision of services of local interest within the network is enabled by the emerging technology of CN micro-clouds. By putting services closer to users, CN micro-clouds pursue not only an improved service performance, but also a low entry barrier for the deployment of alternatives to mainstream Internet services within the CN. Unfortunately, the provisioning of the services is not so simple. Due to the large and irregular topology, high software and hardware diversity and different service requirements in CNs, a "careful" placement of micro-cloud services over the network is required. First, in order to understand the micro-cloud service requirements for a successful operation in CNs, we perform deployment, feasibility analysis and in-depth performance assessment of popular CN micro-cloud services such as distributed storage, live video-streaming and service discovery. We characterize and define workload upper bounds for successful operations of such services and perform cross-layer analysis and optimizations to improve the service performance. This deployment experience supports the feasibility of CN micro-clouds and our measurements contribute to understand the performance of services and applications in this challenging environment. Then, in order to improve the performance of the services on the network level over which a service host provides a service to client nodes, it is necessary to adapt the logical network topology to both external (e.g., wireless connectivity, node availability) and internal (e.g., service copies, service demand) factors. To achieve this, we propose to leverage state information about the network to inform service placement decisions, and to do so through an i) exploratory algorithm PASP (Policy-aware Service Placement) that minimizes the service overlay diameter, while fulfilling service specific criteria and ii) through a fast and low-complexity service placement heuristic BASP (Bandwidth and Availability-aware Service Placement), which maximizes bandwidth between nodes and improves user QoS. Our results show that PASP and BASP consistently outperform the existing in-place strategies in the Guifi.net CN, with respect to bandwidth, availability and latency when used with real CN micro-cloud services. Since this improvement translates in the QoE (Quality of Experience) perceived by the user, our results are relevant for contributing to higher QoE, a crucial parameter for using services from volunteer-based systems. Las redes comunitarias (Community Networks - CNs) han cobrado impulso en los últimos años en respuesta a la creciente demanda de conectividad de red en zonas rurales y urbanas. Estas redes, desplegadas y gestionadas por voluntarios, ofrecen diversos servicios a sus miembros. Si bien el acceso a Internet es el servicio más popular ofrecido a sus participantes, la provisión de servicios de interés local dentro de la red está siendo possible gracias a la tecnología emergente de micro-clouds en CNs. Al acercar los servicios a los usuarios, los micro-clouds persiguen no sólo una mejor experiencia de uso del servicio, sino también bajar la barrera de entrada para el despliegue de alternativas a los servicios de Internet convencionales dentro de la CN. Lamentablemente, la provisión de servicios no es tan simple. Debido a una topología grande e irregular, a la alta diversidad de software y hardware y a los diferentes requisitos de servicio en las CNs, es necesaria una colocación ”cuidadosa” de los servicios de micro-cloud a través de la red. Primero, para comprender los requisitos del micro-cloud para su operación exitosa en CNs, se realiza el despliegue, el análisis de factibilidad y la evaluación en profundidad del rendimiento de servicios populares de micro-cloud tales como almacenamiento distribuido, retransmisión de vídeo y descubrimiento de servicios. Se caracterizan y definen los límites superiores de la carga de trabajo para una operación exitosa de tales servicios y se realizan análisis y optimizaciones entre capas para mejorar el rendimiento del servicio. Esta experiencia de implementación apoya la factibilidad de los micro-clouds comunitarios, y las mediciones mediciones obtenidas contribuyen a comprender el rendimiento de los servicios y aplicaciones en este desafiante entorno. A continuación, con el fin de mejorar el rendimiento de los servicios a nivel de red sobre la que un servidor de servicios proporciona un recurso a los nodos cliente es necesario adaptar la topología de red lógica tanto a factores externos (por ejemplo, conectividad inalámbrica, disponibilidad de nodo) como internos (por ejemplo, copias de servicio, demanda de servicios). Para lograr esto, se propone aprovechar la información de estado de la red para informar de las decisiones de colocación de servicios, y hacerlo a través de un i) algoritmo exploratorio PASP (Policy-aware Service Placement) que minimiza el diámetro del overlay de servicio, cumpliendo Y ii) a través de una heurística de colocación de servicios rápida y de baja complejidad BASP (Bandwidth and Availability-aware Service Placement) que maximiza el ancho de banda entre nodos y mejora la QoS del usuario. Los resultados obtenidos muestran que PASP y BASP consistentemente superan las estrategias existentes en la CN Guifi.net, respecto al ancho de banda, disponibilidad y latencia cuando se usan con servicios de micro-cloud. Dado que esta mejora se traduce en la QoE (calidad de experiencia) percibida por el...

Summary: Decentralization, in the form of mesh networking and blockchain, two promising technologies, is coming to the telecommunications industry. Mesh networking allows wider low‐cost Internet access with infrastructures built from routers contributed by diverse owners, whereas blockchain enables transparency and accountability for investments, revenue, or other forms of economic compensations from sharing of network traffic, content, and services. Crowdsourcing network coverage, combined with crowdfunding costs, can create economically sustainable yet decentralized Internet access. This means that every participant can invest in resources and pay or be paid for usage to recover the costs of network devices and maintenance. While mesh networks and mesh routing protocols enable self‐organized networks that expand organically, cryptocurrencies and smart contracts enable the economic coordination among network providers and consumers. We explore and evaluate two existing blockchain software stacks, Hyperledger Fabric (HLF) and Ethereum geth with Proof of Authority (PoA) intended as a local lightweight distributed ledger, deployed in a real city‐wide production mesh network and in laboratory network. We quantify the performance and bottlenecks and identify the current limitations and opportunities for improvement to serve locally the needs of wireless mesh networks, without the privacy and economic cost of relying on public blockchains.

Summary: Community networks are self‐organized and decentralized communication networks built and operated by citizens, for citizens. The consolidation of today's cloud technologies offers now, for community networks, the possibility to collectively develop community clouds, building upon user‐provided networks and extending toward cloud services. Cloud storage, and in particular secure and reliable cloud storage, could become a key community cloud service to enable end‐user applications. In this paper, we evaluate in a real deployment the performance of Tahoe least‐authority file system (Tahoe‐LAFS), a decentralized storage system with provider‐independent security that guarantees privacy to the users. We evaluate how the Tahoe‐LAFS storage system performs when it is deployed over distributed community cloud nodes in a real community network such as Guifi.net. Furthermore, we evaluate Tahoe‐LAFS in the Microsoft Azure commercial cloud platform, to compare and understand the impact of homogeneous network and hardware resources on the performance of the Tahoe‐LAFS. We observed that the write operation of Tahoe‐LAFS resulted in similar performance when using either the community network cloud or the commercial cloud. However, the read operation achieved better performance in the Azure cloud, where the reading from multiple nodes of Tahoe‐LAFS benefited from the homogeneity of the network and nodes. Our results suggest that Tahoe‐LAFS can run on community network clouds with suitable performance for the needed end‐user experience. Copyright © 2015 John Wiley & Sons, Ltd.