A Cooperative Multi-Agent System for Crowd Sensing Based Estimation in Smart Cities

Abstract

The concept of Smart City has spread as a solution to ensure better access to information and services to citizens, but also as a means to reduce the environmental footprint of cities. To this end, a continuous and wide observation of the environment is necessary to analyze information that enables government bodies to act on the environment appropriately. Moreover, a diffused acquisition of information requires adequate infrastructure and proper devices, which results in relevant installation and maintenance costs. Our proposal enables reducing the number of necessary sensors to be deployed while ensuring that information is available at any time and anywhere. We present the HybridIoT system to cope with the lack of environmental information in the urban context through an estimation technique that integrates heterogeneous data acquired from some different sensors. HybridIoT can be deployed in large-scale contexts and ensures data accessibility even if devices enter or leave the system at any time and everywhere. We compare the results to those obtained by state-of-the-art techniques to assess the validity of our proposal, in particular concerning the properties of openness, large-scale, and heterogeneity, of primary importance in the context of the development of systems to be deployed in the smart city context.

Edge-Based Missing Data Imputation in Large-Scale Environments

“…The use of statistical techniques proves to be efficient to fill the info gaps [14]. In particular, imputation in statistics is the process of replacing missing data with substituted values, thus addressing the pitfalls caused by the presence of missing data [13].…”

“…In particular, imputation in statistics is the process of replacing missing data with substituted values, thus addressing the pitfalls caused by the presence of missing data [13]. This difficulty is mainly related to installation and maintenance costs and the use of statistical techniques proves to be efficient to fill the info gaps [14]. In particular, imputation in statistics is the process of replacing missing data with substituted values, thus addressing the pitfalls caused by the presence of missing data [13].…”

“…The computation is done close to sensors which acquired information from a direct observation of the physical environment. Figure 1 shows the main steps of our methodology, which is based on two types of agents: • type 1 has fixed perceiving devices and it is associated to a Voronoi region which delimits the rough relevance area of the devices [18]; • type 2 has devices not capable of perceiving the environment for some reason and it needs to cooperate with close working devices to provide information [14].…”

Multi-Agent System for Intelligent Urban Traffic Management Using Wireless Sensor Networks Data

“…David Eneko Ruiz de Gauna et al [ 17 ] propose an approach that combines electric vehicles with the advantages of multi-agent systems in order to solve the problems of pollution and congestion. Davide Andrea Guastella et al [ 18 ] designed a cooperative multi-agent system to reduce the number of necessary sensors to be deployed in a smart city.…”

Learn to Sense vs. Sense to Learn: A System Self-Integration Approach

“…Thanks to the diffusion of IoT devices, sensors-based applications are spreading in different fields such as bioprocess monitoring [4], aerospace [1], missing data imputation [2]. Such applications mainly aggregate data acquired from multiple IoT devices.…”