Seguridad y calidad de los datos en redes de sensores inalámbricas IoT

Abstract

In recent years, there are several emerging technologies that have made their way among others already consolidated, such as the Internet of Things (IoT), Blockchain or Big Data, among others; and with applications in various sectors, such as smart homes, smart cities, the agri-food sector where applications are being developed in the area of smart farming or food traceability, production and logistics chains, e-health, defence, vehicle safety, even in industry, among others. In many cases, these technologies are being used jointly in order to take advantage of the characteristics of each of them to address some of the challenges involved in deploying this type of technology while ensuring the security conditions and the quality of the information processed by the systems. The context of the research focuses on the primary sector, specifically in the agricultural sector, in the use of IoT technology, especially IoT sensors interconnected through wireless networks in order to monitor environmental variables of livestock, and ultimately of any element of the beginning of the production chain developing methods for communication between sensors and the rest of the system is done securely, that the data collected through IoT sensors become part of a blockchain to take advantage of the properties of this technology, previously deploying a data curation process through a Big Data pipeline that ensures the quality of the data to be introduced into the business logic and as a result favour subsequent decision making with the greatest possible guarantee. To this end, firstly, a research work has been carried out on the guarantees offered by the range of blockchain technologies in terms of security and integrity through a systematic literature review from the perspective of privacy and anonymisation, which provides us with a panoramic view of the characteristics of this technology and the strong relationship between privacy and anonymisation in most of the fields of application of blockchain. It can be observed that there are different degrees of privacy enforcement depending on the techniques used to implement anonymisation, with traceability of transactions being one of the privacy risks. It then addresses the challenge of ensuring the integrity of information in the IoT sensor network when transmitting data between IoT devices. By using blockchain technology, the integrity of data transactions between entities is effectively guaranteed, eliminating any possibility of unauthorised access to the wireless sensor network and data injection by malicious devices, but this does not prevent a sensor from entering erroneous data. Consequently, a robust mechanism that is based on smart contracts and blockchain technology has been designed for the reliable processing of data collected through IoT sensors. In a last phase, and solving the aforementioned problem about the validity of the data collected and sent by the sensors, we focus on evaluating the quality of the data collected by the sensors and sent to the blockchain, in different scenarios (offline and online), so that we can determine whether each data entered into the system through the IoT sensors is susceptible to be used, or on the contrary if it is discarded for not reaching sufficient quality. For this purpose, during this thesis, an IoT-based Big Data pipeline has been designed that integrates data transformation and integration tools as well as a configurable decision model based on Decision Model Notation (DMN) to evaluate the quality of the data. The solution provided in this thesis allows us to measure the quality of the data and thus incorporate into the system only those that exceed a previously established minimum. In this way, we avoid serious inaccuracies that involve the whole decision-making process, or the obtaining of non-optimal decisions because it has been considered that a data may be complete but not valid due to the possible malfunctioning of a sensor node.