Long-Term Data Traffic Forecasting for Network Dimensioning in LTE with Short Time Series

Network dimensioning is a critical task in current mobile networks, as any failure in this process leads to degraded user experience or unnecessary upgrades of network resources. For this purpose, radio planning tools often predict monthly busy-hour data traffic to detect capacity bottlenecks in advance. Supervised Learning (SL) arises as a promising solution to improve predictions obtained with legacy approaches. Previous works have shown that deep learning outperforms classical time series analysis when predicting data traffic in cellular networks in the short term (seconds/minutes) and medium term (hours/days) from long historical data series. However, long-term forecasting (several months horizon) performed in radio planning tools relies on short and noisy time series, thus requiring a separate analysis. In this work, we present the first study comparing SL and time series analysis approaches to predict monthly busy-hour data traffic on a cell basis in a live LTE network. To this end, an extensive dataset is collected, comprising data traffic per cell for a whole country during 30 months. The considered methods include Random Forest, different Neural Networks, Support Vector Regression, Seasonal Auto Regressive Integrated Moving Average and Additive Holt–Winters. Results show that SL models outperform time series approaches, while reducing data storage capacity requirements. More importantly, unlike in short-term and medium-term traffic forecasting, non-deep SL approaches are competitive with deep learning while being more computationally efficient.

On Mathematical Modelling of Automated Coverage Optimization in Wireless 5G and beyond Deployments

The need to optimize the deployment and maintenance costs for service delivery in wireless networks is an essential task for each service provider. The goal of this paper was to optimize the number of service centres (gNodeB) to cover selected customer locations based on the given requirements. This optimization need is especially emerging in emerging 5G and beyond cellular systems that are characterized by a large number of simultaneously connected devices, which is typically difficult to handle by the existing wireless systems. Currently, the network infrastructure planning tools used in the industry include Atoll Radio Planning Tool, RadioPlanner and others. These tools do not provide an automatic selection of a deployment position for specific gNodeB nodes in a given area with defined requirements. To design a network with those tools, a great deal of manual tasks that could be reduced by more sophisticated solutions are required. For that reason, our goal here and our main contribution of this paper were the development of new mathematical models that fit the currently emerging scenarios of wireless network deployment and maintenance. Next, we also provide the design and implementation of a verification methodology for these models through provided simulations. For the performance evaluation of the models, we utilize test datasets and discuss a case study scenario from a selected district in Central Europe.

Design, Implementation, and Empirical Validation of an IoT Smart Irrigation System for Fog Computing Applications Based on LoRa and LoRaWAN Sensor Nodes

Climate change is driving new solutions to manage water more efficiently. Such solutions involve the development of smart irrigation systems where Internet of Things (IoT) nodes are deployed throughout large areas. In addition, in the mentioned areas, wireless communications can be difficult due to the presence of obstacles and metallic objects that block electromagnetic wave propagation totally or partially. This article details the development of a smart irrigation system able to cover large urban areas thanks to the use of Low-Power Wide-Area Network (LPWAN) sensor nodes based on LoRa and LoRaWAN. IoT nodes collect soil temperature/moisture and air temperature data, and control water supply autonomously, either by making use of fog computing gateways or by relying on remote commands sent from a cloud. Since the selection of IoT node and gateway locations is essential to have good connectivity and to reduce energy consumption, this article uses an in-house 3D-ray launching radio-planning tool to determine the best locations in real scenarios. Specifically, this paper provides details on the modeling of a university campus, which includes elements like buildings, roads, green areas, or vehicles. In such a scenario, simulations and empirical measurements were performed for two different testbeds: a LoRaWAN testbed that operates at 868 MHz and a testbed based on LoRa with 433 MHz transceivers. All the measurements agree with the simulation results, showing the impact of shadowing effects and material features (e.g., permittivity, conductivity) in the electromagnetic propagation of near-ground and underground LoRaWAN communications. Higher RF power levels are observed for 433 MHz due to the higher transmitted power level and the lower radio propagation losses, and even in the worst gateway location, the received power level is higher than the sensitivity threshold (−148 dBm). Regarding water consumption, the provided estimations indicate that the proposed smart irrigation system is able to reduce roughly 23% of the amount of used water just by considering weather forecasts. The obtained results provide useful guidelines for future smart irrigation developers and show the radio planning tool accuracy, which allows for optimizing the sensor network topology and the overall performance of the network in terms of coverage, cost, and energy consumption.

Design and Empirical Validation of a LoRaWAN IoT Smart Irrigation System

In some parts of the world, climate change has led to periods of drought that require managing efficiently the scarce water and energy resources. This paper proposes an IoT smart irrigation system specifically designed for urban areas where remote IoT devices have no direct access to the Internet or to the electrical grid, and where wireless communications are difficult due to the existence of long distances and multiple obstacles. To tackle such issues, this paper proposes a LoRaWAN-based architecture that provides long distance and communications with reduced power consumption. Specifically, the proposed system consists of IoT nodes that collect sensor data and send them to local fog computing nodes or to a remote cloud, which determine an irrigation schedule that considers factors such as the weather forecast or the moist detected by nearby nodes. It is essential to deploy the IoT nodes in locations within the provided coverage range and that guarantee good speed rates and reduced energy consumption. Due to this reason, this paper describes the use of an in-house 3D-ray launching radio-planning tool to determine the best locations for IoT nodes on a real medium-scale scenario (a university campus) that was modeled with precision, including obstacles such as buildings, vegetation, or vehicles. The obtained simulation results were compared with empirical measurements to assess the operating conditions and the radio planning tool accuracy. Thus, it is possible to optimize the wireless network topology and the overall performance of the network in terms of coverage, cost, and energy consumption.