Multi-dimensional Urban Sensing Using Crowdsensing Data

Description

Chaocan Xiang is an Associate Professor at the College of Computer Science, Chongqing University, China. He received his bachelor’s degree and Ph.D. from Nanjing Institute of Communication Engineering, China, in 2009 and 2014, respectively. He subsequently studied at the University of Michigan-Ann Arbor in 2017 (supervised by Prof. Kang G. Shin, IEEE Life Fellow, ACM Fellow). His research interests mainly include UAVs/vehicle-based crowdsensing, urban computing, Internet of Things, Artificial Intelligence, and big data. He has published more than 50 papers, including over 20 in leading venues such as IEEE Transactions on Mobile Computing, IEEE Transactions on Parallel and Distributed Systems, IEEE INFOCOM, and ACM Ubicomp. He has received a best paper award and a best poster award at two international conferences. Panlong Yang is a full Professor at the University of Science and Technology of China. He has been supported by the NSF Jiangsu through a Distinguished Young Scholarship and was honored as a CCF Distinguished Lecturer in 2015. He has published over 150 papers, including 40 in CCF Class A. Since 2012, he has supervised 14 master’s and Ph.D. candidates, including two excellent dissertation winners in Jiangsu Province and the PLA education system. He has been supported by the National Key Development Project and NSFC projects. He has nominated by ACM MobiCom 2009 for the best demo honored mention awards, and won best paper awards at the IEEE MSN and MASS. He has served as general chair of BigCom and TPC chair of IEEE MSN. In addition, he has served as a TPC member of INFOCOM (CCF Class A) and an associate editor of the Journal of Communication of China. He is a Senior Member of the IEEE (2019). Fu Xiao received his Ph.D. in Computer Science and Technology from the Nanjing University of Science and Technology, Nanjing, China, in 2007. He is currently a Professor and Dean of the School of Computer, Nanjing University of Posts and Telecommunications. He has authored more than 60 papers in respected conference proceedings and journals, including IEEE INFOCOM, ACM Mobihoc, IEEE JASC, IEEE/ACM ToN, IEEE TPDS, IEEE TMC, etc. His main research interest is in the Internet of Things. He is a member of the IEEE Computer Society and the Association for Computing Machinery. Xiaochen Fan received his B.S. degree in Computer Science from Beijing Institute of Technology, Beijing, China, in 2013, and his Ph.D. from the University of Technology Sydney, NSW, Australia, in 2021. His research interests include mobile/pervasive computing, deep learning, and Internet of Things (IoT). He has published over 25 peer-reviewed papers in high-quality journals and IEEE/ACM international conference proceedings. Preface Acknowledgments Contents Part I: How to Collect Crowdsensing Data (Multi-dimensional Fundamental Issues) Chapter 1: Incentivizing Platform-Users with Win-Win Effects 1.1 Introduction 1.2 Related Work 1.3 System Model and Problem Formulation 1.3.1 System Model 1.3.2 Example of Personalized Bidding Scenario 1.3.3 Problem Formalization 1.3.4 Analysis of Problem Complexity 1.4 Design of Picasso 1.4.1 Bid Description in 3-D Space PB Description in 3-D Space Formal Framework of Bid Description Using 3-D Expressive Space PB Description Method Theoretical Analysis 1.4.2 Task Allocation Based on Dependency Graph Construction of Task Dependency Graph PB Decomposition for Efficient Task Allocation Problem Transformation by Decomposing the Task Dependency Graph Task Allocation with Constant-Factor Approximation Theoretical Analysis PB Recombination for Strategy-Proof Payment Truthful Payment Scheme for Non-PB Based on Critical Prices Truthful Payment Scheme for PB Based on Graph Recombination Theoretical Analysis 1.5 Performance Evaluation of Picasso 1.5.1 Simulations Simulation Methodology and Settings Results 1.5.2 Trace-Driven Case Study of Gigwalk Evaluation Methodology and Settings Results 1.6 Discussion and Future Work 1.7 Conclusion References Chapter 2: Task Recommendation Based on Big Data Analysis 2.1 Introduction 2.2 Related Work 2.3 Motivation 2.3.1 Crowdsourcing-Based User Studies 2.3.2 Large-Scale Dataset Collection and Analysis 2.4 LSTRec Design for MOVE-CS 2.4.1 Model Design 2.4.2 Research Problem and Challenge Analysis 2.5 Key Algorithm Design for LSTRec 2.5.1 Pick-Up Profit Heatmap Construction 2.5.2 Differentiation-Aware Sensing Reward Design 2.5.3 Submodularity-Based Task Recommendation Algorithm 2.6 Evaluation 2.6.1 Emulation Methodology and Settings 2.6.2 Results of Model Evaluation 2.6.3 Results of Algorithm Evaluation 2.7 Conclusion References Chapter 3: Data Transmission Empowered by Edge Computing 3.1 Introduction 3.2 Related Work 3.3 Motivation 3.3.1 Uncovering Missing Data Issue in Large-Scale ITSs 3.3.2 Experimental Explorations of Spatio-Temporal Correlations on Traffic Data 3.3.3 Implementation Dilemmas for Large-Scale Traffic Recovery 3.4 System Model and Problem Formulation 3.4.1 System Model of Edge Computing 3.4.2 Problem Formalization 3.5 System Design 3.5.1 Sub-optimal Deployment of Edge Nodes Problem Reformulation Local Search-Based Suboptimal Deployment 3.5.2 Accurate Traffic Data Recovery Based on Low-Rank Theory Experimental Analysis of Low-Rank Accurate Traffic Recovery Based on Low-Rank Theory 3.6 Traces-Based Evaluations 3.6.1 Experimental Methodology and Settings Dataset and Experimental Methodology Baseline Methods and Evaluation Metrics 3.6.2 Experimental Results Evaluations of Edge Node Deployment Evaluations of Traffic Recovery 3.7 Discussion and Future Work 3.8 Conclusion References Part II: How to Use Crowdsensing Data for Smart Cities (Multi-dimensional Applications) Chapter 4: Environmental Protection Application: Urban Pollution Monitoring 4.1 Introduction 4.2 Related Work 4.3 System Model and Problem Description 4.3.1 System Model 4.3.2 Problem Description 4.4 Iterative Truthful-Source Identification Algorithm 4.4.1 Algorithm Design Truthful Probability Estimation Sensor Efficiency Estimation 4.4.2 Algorithm Description and Analysis 4.5 Experimental Evaluations 4.5.1 Simulation Methodology and Settings 4.5.2 Performance Evaluations Algorithm Convergence Algorithm Robustness Performance Impact of the Number of Users 4.6 Discussion 4.7 Conclusion References Chapter 5: Urban Traffic Application: Traffic Volume Prediction 5.1 Introduction 5.2 Related Work 5.2.1 Reusing Building Data 5.2.2 Traffic Volume Prediction 5.2.3 Traffic Prediction Models 5.3 System Overview 5.4 Building-Traffic Correlation Analysis with Multi-Source Datasets 5.4.1 Correlation Analysis with Building Occupancy Data Correlation Quantification with Metrics Correlation Verification with Google Maps 5.4.2 Correlation Analysis with Environmental Data 5.5 Accurate Traffic Prediction with Cross-Domain Learning of Building Data 5.5.1 Problem Formulation 5.5.2 Attention Mechanisms-Based Encoder-Decoder Recurrent Neural Network Encoder with Cross-Domain Attention Occupancy Component Environmental Component Decoder with Temporal Attention 5.6 Performance Evaluation 5.6.1 Experimental Methodology and Settings Dataset Description Baseline Methods Evaluation Metrics and Parameter Settings 5.6.2 Experimental Evaluations Evaluations on Overall Prediction Results Comparisons with Baselines Evaluation of Parameters Evaluation of Cross-Domain Learning Evaluation of Attention Weight Extensive Evaluations of Comparison Baseline Methods with the Weekday´s Data and the Weekend´s Data Comparison with State-of-the-Art Methods Based on Different Data Sources 5.7 Discussion 5.7.1 Applicable Conditions of BuildSenSys 5.7.2 Sensing Coverage of BuildSenSys 5.7.3 Extension to Large-Scale Scenarios 5.7.4 Privacy and Security of Building Data 5.8 Conclusion References Chapter 6: Airborne Sensing Application: Reusing Delivery Drones 6.1 Introduction 6.2 Related Work 6.2.1 Crowdsensing Based on Other Mobile Devices 6.2.2 Crowdsensing Based on Drones Summary 6.3 Models and Problem Formulation 6.3.1 Delivering and Sensing Models for Delivery Drones 6.3.2 Energy Consumption Model for Delivery Drones On-Site Experimental Explorations Energy Consumption Model Model Analysis 6.3.3 Problem Formulation RT Problem RTW Problem Objective and Constraints Challenges 6.4 Solution Design 6.4.1 Overview 6.4.2 Route-Time Joint Allocation Algorithm Equivalent Objective Function Construction Approximation Algorithm for Route-Time Joint Allocation 6.4.3 Route-Time-Weight Joint Optimization Algorithm 6.4.4 Theoretical Analysis of Algorithms Analysis of Objective Function and Constraints Analysis of Algorithm Performance 6.5 Experimental Evaluations 6.5.1 Traces-Based Simulations Evaluation Methodology and Settings Performances of RT-alg Algorithm Performances of RTW-Algorithm Case Study of Drone-Based AQI Monitoring 6.5.2 Field Experiments System Implementation Experimental Settings and Baselines Experimental Results Summary 6.6 Discussions and Future Works 6.6.1 Generalizing Drone Route Model 6.6.2 Incentives Behind Drone Sharing 6.6.3 Impacts of Delivery Delay 6.6.4 Factors in Sensing Utility 6.6.5 Impact of Sensors and Battery Weight 6.7 Conclusion References Part III: Open Issues and Conclusions Chapter 7: Open Issues 7.1 Exploring New Urban Applications 7.2 Utilizing More Crowdsensing Data 7.3 Concerning Privacy Protection References Chapter 8: Conclusions