Wi-Fi based city users’ behaviour analysis for smart city

Abstract

a b s t r a c tMonitoring, understanding and predicting city user behaviour (hottest places, trajectories, flows, etc.) is one the major topics in the context of Smart City management. People flow surveillance provides valuable information about city conditions, useful not only for monitoring and controlling the environmental conditions, but also to optimize the deliverying of city services (security, clean, transport,..). In this context, it is mandatory to develop methods and tools for assessing people behaviour in the city. This paper presents a methodology to instrument the city via the placement of Wi-Fi Access Points, AP, and to use them as sensors to capture and understand city user behaviour with a significant precision rate (the understanding of city user behaviour is concretized with the computing of heat-maps, origin destination matrices and predicting user density). The first issue is the positioning of Wi-Fi AP in the city, thus a comparative analyses have been conducted with respect to the real data (i.e., cab traces) of the city of San Francisco. Several different positioning methodologies of APs have been proposed and compared, to minimize the cost of AP installation with the aim of producing the best origin destination matrices. In a second phase, the methodology was adopted to select suitable AP in the city of Florence (Italy), with the aim of observing city users behaviour. The obtained instrumented Firenze Wi-Fi network collected data for 6 months. The data has been analysed with data mining techniques to infer similarity patterns in AP area and related time series. The resulting model has been validated and used for predicting the number of AP accesses that is also related to number of city users. The research work described in this paper has been conducted in the scope of the EC funded Horizon 2020 project Resolute ( http://www.resolute-eu.org ), for early warning and city resilience.

Data Mining and Machine Learning to Promote Smart Cities: A Systematic Review from 2000 to 2018

“…Yuan et al [22] T-Drive: Enhancing Driving Directions with Taxi Drivers' Intelligence Rudin et al [23] Machine Learning for the New York City Power Grid Jurado et al [24] Hybrid methodologies for electricity load forecasting: Entropy-based feature selection with machine learning and soft computing techniques Pérez-Chacón et al [25] Big data analytics for discovering electricity consumption patterns in smart cities Peña et al [26] Rule-based system to detect energy efficiency anomalies in smart buildings, a data mining approach Liu et al [27] A machine learning-based method for the large-scale evaluation of the qualities of the urban environment Muhammed et al [28] UbeHealth: A Personalized Ubiquitous Cloud and Edge-Enabled Networked Healthcare System for Smart Cities Massana et al [29] Identifying services for short-term load forecasting using data driven models in a Smart city platform Wang et al [30] Identification of key energy efficiency drivers through global city benchmarking: a data driven approach Abbasi and El Hanandeh [31] Forecasting municipal solid waste generation using artificial intelligence modelling approaches Badii et al [32] Predicting Available Parking Slots on Critical and Regular Services by Exploiting a Range of Open Data Madu et al [33] Urban sustainability management: A deep learning perspective Gomede et al [34] Application of Computational Intelligence to Improve Education in Smart Cities. Cramer et al [35] An extensive evaluation of seven machine learning methods for rainfall prediction in weather derivatives You and Yang [36] Urban expansion in 30 megacities of China: categorizing the driving force profiles to inform the urbanization policy Nagy and Simon [37] Survey on traffic prediction in smart cities Belhajem et al [38] Improving Vehicle Localization in a Smart City with Low Cost Sensor Networks and Support Vector Machines Fernández-Ares et al [39] Studying real traffic and mobility scenarios for a Smart City using a new monitoring and tracking system Belhajem et al [40] Improving low cost sensor based vehicle positioning with Machine Learning Gopalakrishnan [41] Deep Learning in Data-Driven Pavement Image Analysis and Automated Distress Detection: A Review Khan et al [42] Smart City and Smart Tourism: A Case of Dubai Idowu et al [43] Applied machine learning: Forecasting heat load in district heating system Bellini et al [44] Wi-Fi based...…”

“…Bellini et al [44] developed a methodology to evaluate people's behavior in the city from Wi-Fi access points. Cluster analyses were applied to extract the most frequented places and typical city users' behavior, among other data.…”

“…The studies also exhibit that the proposed studies are motivated to find solutions for specific domains of the city (e.g., traffic, mobility, environment, and so on). Clustering analyses were the most used approaches in the studies [22,25,30,39,44,46,50,52,54,57] and later, supported vector machine [23,31,38,40,43,49,57].…”

Analyzing Public Transportation Offer wrt Mobility Demand

“…For example, in cities in which the present of Tourists is very relevant (e.g., Venezia, Roma, Firenze) which cannot be neglected with respect to the citizens, commuters, and students. In this case, data that come from the cellular networks [10] and Wi-Fi [11], considering, respectively, access points and regions in the city, can help in analyzing different aspects (e.g., how many of them are daily present in the city, how long they stay, where they go/come in long term/distance) by investigating people's flow.…”

Real-Time Pedestrian Flow Analysis Using Networked Sensors for a Smart Subway System

“…In particular, radio-frequency (RF) signal-based technologies have shown high accuracy at relatively low cost compared to other technologies. Therefore, in many indoor measurement studies, RF signal-based wireless technologies such as Wi-Fi positioning system (WPS) have been recently used [7]. A WPS utilizes the 2.4 GHz band signal of the IEEE 802.xx standard, and is the most widely used IPS technology at present.…”