Introduction: Mobile Data Management
Mobile data management, which our lab focuses on, mainly includes mobile database techniques, and moving objects data management. Mobile database techniques include mobile transaction management, data caching and replication, synchronization and publication, etc. Moving objects data management includes modeling and tracking of dynamic location information, uncertainty management, spatio-temporal data access languages, indexing and scalability issues, location-based query processing, data mining (including traffic and location prediction), location dissemination, location privacy preserving, location fusion and synchronization.
[Top]        
Motivation
The combination of computing techniques and wireless networks makes mobile computing more and more pervasive. Compared with traditional distributed computing environment based on stable networks, mobile computing has the following features: mobility, frequent disconnection, variety of bandwidth, asymmetry of network communicating, scalability, limited power of mobile devices, low reliability of the networks, and so on. In such environments, many new applications deal with a significant amount of data leading to the need of mobile data management techniques.
[Top]        
Research work
  • Mobile Database System
  • There has been much work in the area of mobile database research. In mobile database systems, new features such as mobility, disconnection and long-lived transactions make traditional transaction processing schemes no longer suited. To solve this problem, a mobile transaction model, O2PC-MT, is proposed in [1]. By combining Optimistic Concurrency Control and Two-Phase Commit, O2PC-MT provides a flexible and effective support for mobile transaction processing. Mobility of users also has a significant impact on data replication. As a result, the various replica control protocols that exist today in traditional distributed and multi-database environments are not applicable any more. A novel mobile database replication scheme, the Transaction-Level Result-Set Propagation (TLRSP) model, is proposed in [2, 3]. In TLRSP model, mobile users are allowed to access local replicas of the database and submit local transactions when system is disconnected. A conflict detection and resolution strategy based on TLRSP is discussed in [3], using which locally committed transactions could be sent to the fixed database server for conflict reconciliation and result-set incorporation when system is reconnected. Based on above approaches, a prototype, KingBase Lite, is proposed, which is an embedding mobile database system applied in the mobile devices such as PDA, palm computer and mobile phone.


  • Footprint Database
  • Very few work has been done to deal with small footprint database. Because of high speed of reading, low cost of power and other benefits, flash memory is often used in small intelligent devices. Several flash-based storage models, such as Flat-Storage (FS), Domain-Storage (DS), Ring-Storage (RS) and ID based Storage models, have been proposed. Query processing and optimization is designed based on these models. Recovery technique is considered in [4], which summarizes the main problems of flash-based small DBMS and proposes a log-structured data storage and index model that can support efficient recovery.


  • Modeling for Moving Objects
  • Plenty of work has been done in different aspects of moving objects data management. Early work focuses on moving objects management in free space. A discrete spatio-temporal trajectory based model in [5] uses trajectories to represent dynamic attributes of moving objects, which can submit moving plans and set up different threshold to trigger the update of its future trajectory and corresponding index. We have an observation that objects usually move within a constrained networks in the real world, e.g. vehicles move in road networks. Therefore we propose a new model based on graph, called GCA (Graph Cellular Automata)[6], which combines the information of road network and the movement of objects in the road and makes the location management more efficiently. [6] proposes a simulation-based prediction method based on the model, which provides higher accuracy than the linear prediction method when applied to real traffic condition. Tracking and update problems are also the interest of point. A simulation-based prediction model in [7] provides a group update policy, which puts the objects in groups and only the central object in each group reports update. This model brings more accurate location prediction for objects movements in a traffic road network while lowers the update frequency and assures the location precision.


  • Indexing for Moving Objects in Spatial Networks
  • Development of efficient indexes is a challenge due to frequent object movement. A novel spatio-temporal index is proposed in [8], which is based on PMR quadtree and adopts a trajectory segment shared structure while depicting an efficient update algorithm. Considering the influence of road network to the movement of moving objects and their inherent impacts among themselves, we proposed a new indexing--ANR-Tree (Adaptive Network R-tree) [9, 10], which employs a dynamic structure called AU (Adaptive Unit) to expand R-tree, thus can support frequent moving object updates with less cost. AU groups neighboring objects with similar movement patterns. To reduce updates, a AU captures the movement bounds of the objects based on a prediction method, which considers the road-network constraints and stochastic traffic behavior. A spatial index for the road network is then built over the adaptive unit structures. The ANR-tree extends the R-tree with adaptive units. The predicted movement of the adaptive unit is given by two trajectory bounds based on different assumptions on the traffic conditions and obtained from the simulation.


  • Querying for Moving Objects
  • Nearest neighbor search is one of the fundamental problems in the field of spatial databases. A k-NN query computes the k data objects that lie closest to a given query point. In the real world applications,severs often receive a lot of query requests, so how to process these queries efficiently and achieve better I/O efficiency is a challenging goal. We study the multiple KNN queries processing for moving objects in road networks and cluster the query points and propose a Clu-MQN (Clustering based Multiple Queries in road Network)[11] algorithm, which processes the queries in the same cluster synchronously, so the efficiency of query processing is greatly improved.

    Two important applications of location data mining are clustering and density query, which can find potential traffic congestion and utilize traffic management or prediction.


  • Density Queries for Moving Objects
  • A density query returns the regions with a density higher than some user-specified threshold. Identifying dense regions is valuable for many applications like traffic control, resource scheduling, and collision probability evaluation. We provide a definition of continuous density queries for moving objects, which returns useful answers and is amenable to efficient computation[12]. Furthermore, we propose the notion of safe interval for dense/spare regions to support efficient processing of continuous density queries. We also define an effective density query for moving objects in road network [13]. We use a two-phase algorithm to identify the dense areas based on the summary information of the cluster units.


  • Clustering for Moving Objects in Spatial Networks
  • For some new applications, real-time data analysis such as clustering analysis is becoming one of the most important requirements, especially, clustering moving objects in Spatial Networks. One of objectives for clustering objects is to find traffic congestion condition in spatial networks. A unified framework for "clustering moving objects in Spatial Networks" (CMON for short) is proposed in [14]. The goals are to optimize the cost of clustering moving objects and support multiple types of clusters in a single application. The framework is composed of two components: (1) The continuous maintenance of cluster blocks (CBs); (2) The periodical construction of cluster with different criteria based on CBs. A CB groups a set of objects on a road segment in close proximity to each other at present and in the near future. In general, a CB can satisfy two basic requirements: (1) it is inexpensive to maintain in a spatial network setting; and (2) it is able to serve as a building block of different types of application-level cluster. An incremental CB maintenance (including split and merge) algorithms are developed by analyzing the object movement features on a spatial network. Furthermore, an efficient algorithm is presented to periodically construct three kinds of clusters based on CBs. The network features are exploited to reduce the search space and avoid unnecessary computation of network distance.


  • Location-Based Privacy Protection
  • Protection of user’s privacy has been a central issue for location-based services (LBSs). In [16], two kinds of privacy protection requirements in LBS are identified: location anonymity and identifier anonymity. While the location cloaking technique under the k-anonymity model can provide a good protection of users’ privacy, it reduces the resolution of location information and, hence, may degrade the quality of service (QoS). To strike a balance between the location privacy and Qos, a quality-aware anonymity model for protecting location privacy while meeting user specified Qos requirements is presented. In the model, a mobile user can specify the minimum anonymity level requirement upon location privacy as well as the maximum cloaking latency and the maximum cloaking region size requirements upon QoS. In accordance with the model, an efficient directed-graph based cloaking algorithm is developed to achieve both high-quality location anonymity and identifier anonymity.

    There are two types of privacy concerns in location-based services: location privacy and query privacy. Existing work, based on location k-anonymity, mainly focused on location privacy and are insufficient to protect query privacy. In particular, due to lack of semantics, location k-anonymity suffers from query homogeneity attack. In [17], we introduce p-sensitivity, a novel privacyprotection model that considers query diversity and semantic information in anonymizing user locations. We propose a PE-Tree for implementing the p-sensitivity model. Search algorithms and heuristics are developed to efficiently find the optimal p-sensitivity anonymization in the tree. Preliminary experiments show that p-sensitivity provides highquality services without compromising users’ query privacy.

    A lot of location cloaking approaches have been proposed for protecting the location privacy of mobile users. But most of existing works only consider the snapshot location privacy preserving. In [18], we first present location-dependent attack resulting from continuous and dependent location updates. We proposed an incremental clique-based cloaking algorithm, called ICliqueCloak, to defend against location-dependent attack. The main idea is to incrementally maintain maximal cliques for location cloaking in an un-directed graph that takes into consideration the effect of continuous location updates.

    [Top]        
    Systems

    The Kingbase Lite is a three-tier architecture, which is designed for small mobile devices with limited resources, such as palm computer, PDA, mobile phones, and so on. The data synchronization mechanism between the client and the server is based on the optimistic asynchronous replication technique, which can ensure the caching data in the clients and the data in the server are consistent. In view of the limitation of the resource and the network bandwidth, the system also does some optimization on the storage of the caching data and the transmission of the refresh data.


    Phone DB is a small footprint DBMS for mobile phones architecture. The key techniques of Phone DB include: 1) key-based data model (key representation like BerkeleyDB application-defined scheme); 2) write buffer; 3) log-based recovery; 4) log-based record storage.

    [Top]        
    Future work

    In the future, for the moving objects management, the constrained network environments will be much more concerned since in most real life applications, objects move mostly within constrained networks, especially the transportation networks. The main challenges for managing network-constrained moving objects are: data representation and modeling of moving objects in road network, indexing for network-constrained moving objects, efficient NN queries and continuous queries on road network, intelligent analysis of objects movement in transportations, and so on.


    Besides these issues, as we know, the strong growth in wireless communications and the ever increasing availability of mobile multi-purpose devices have created a global computing environment. People communicate, work, and confer using a wide range of devices all connected via an array of communication networks that provide voice and data access regardless of geographic position. This infrastructure aggregation presents a number of challenges especially when it comes to data-intensive applications such as LBS, PIM and applications with sensor network. So, non-traditional issues including semantics of data, location-centric data services, broadcast and multicast delivery, data availability techniques, security of data, as well as privacy questions should be take much care of.

    Mobile computing is one aspect of the pervasive computing, so widely we plan to do some research on the probabilistic data management for pervasive computing. The wide deployment of wireless sensor and RFID (Radio Frequency Identification) devices is one of the key enablers for next-generation pervasive computing applications, including large-scale environmental monitoring and control, context-aware computing, and “smart digital homes”. Sensory readings are inherently unreliable and typically exhibit strong temporal and spatial correlations; effective reasoning over such unreliable streams introduces a host of new data management challenges. Due to the importance of those applications and the rapidly increasing amount of uncertain data collected and accumulated, analyzing large collections of uncertain data has become an important task and has attracted more and more interest from the database community. Recently, uncertain data management has become an emerging hot area in database research and development.

    [Top]        
    Grant
    • 2006-2008 Network-Constrained Moving Objects Management (Principle Investigator)
      Granted by the Natural Science Foundation of China(NSFC) under grant number 60573091
    • 2006-2007 Network-Constrained Moving Objects Management (Principle Investigator)
      Granted by China-France PRA project
    • 2005 Road Network Moving Objects Management (Principle Investigator)
      Granted by CNRS/NSFC
    • 2002-2003 Moving Objects Databases (Principle Investigator)
      Granted by the Key Project of Chinese Ministry of Education (No.03044) and
      the Excellent Young Teachers Program of M0E,P.R.C (EYTP)
    • 1999-2003 Embedded Mobile Database System (Principle Investigator)
      Granted by 863 project and Natural Science Foundation of China(NSFC) under grant number 60073014
    [Top]        
    Patents
    • System and method for moving objects tracking based on group.
      Applied number 200810056097.2(Pending)
    • System for monitoring congestion of road network based on clustering.
      Applied number 200810056097.1(Pending)
    [Top]        
    Publication
    • J. Wen, X. Meng, X. Hao and J. Xu: An Efficient Approach for Continuous Density Queries. Frontiers of Computer Science. Frontiers of Computer Science. Vol. 5(6), Pages:581-595, 2012
    • Z. Huo, X. Meng, R. Zhang: Feel Free to Check In: Privacy Alert against Hidden Location Inference Attacks in GeoSN. Accepted for publication in the 18th International Conference on Database Systems for Advanced Applications (DASFAA 2013). April 22-25, 2013, Wuhan, China. (Regular paper)
    • Z. Huo, X. Meng, J. Xu: Privacy-preserving Query Processing in Cloud Computing. Journal of Frontiers of Computer Science and Technology. Vol 6(5): 385-396, 2012, 6.
    • Z. Huo, X. Meng, H. Hu, Y. Huang: You Can Walk Alone: Trajectory Privacy-Preserving through Significant Stays Protection. In Proceedings of the 17th International Conference of Database Systems for Advanced Applications (DASFAA 2012), pages: 351-366, April 15-19, 2012, Busan, South Korea.
    • J. Zhang, J. Wen, X. Meng: Multi-Tag Route Query Based on Order Constraints in Road Networks. Chinese Journal of Computer, Vol 35(11): 2317-2326, 2012, 11. (NDBC2012, Hefei)(the Sa Shi Xuan Best Paper Award)
    • J. Zhang, X. Meng, X. Zhou, D. Liu: Co-spatial Searcher: Efficient Tag-Based Collaborative Spatial Search on Geo-social Network. In Proceedings of the 17th International Conference of Database Systems for Advanced Applications (DASFAA 2012), pages: 560-575, April 15-19, 2012, Busan, South Korea.
    • J. Zhang, X. Meng: Mobile Web search. Journal of Software. Vol.23 (01): 46-64, 2012.
    • X. Pan, J. Xu, X. Meng: Protecting Location Privacy Against Location-dependent Attacks in Mobile Services. IEEE Transaction on Knowledge and Data Engineering (TKDE).24(8):1506-1519,2012 .(Regular paper)
    • Y. Huang, Z. Huo, X. Meng: CoPrivacy: A Collaborative Location Privacy-preserving Method without Cloaking Region. Chinese Journal of Computers, Vol 34(10):1820-1830, 2011,10.(NDBC2011, Shanghai)(the Sa Shi Xuan Best Paper Award)
    • Z. Huo, X. Meng: A Survey of Trajectory Privacy-Preserving Techniques. Chinese Journal of Computers, Vol 34(10):1820-1830, 2011,10.
    • C. Zhou, X. Meng, Y. Chen: Out-of-order Durable Event Processing in Integrated Wireless Networks. Journal of Pervasive and Mobile Computing (PMCJ). Vol.7,NO.5:595-610,Oct.2011.
    • J. Zhang, D. Liu, X. Meng: Preference-based top-k Spatial Keyword Queries. In Proceedings of the 1st International Workshop on Mobile Location-Based Service (MLBS 2011, in conjunction with Ubicomp). September 17-21, Beijing, 2011: 31-39.
    • Z. Huo, Y. Huang, X. Meng: History Trajectory Privacy-preserving through Graph Partition. In Proceedings of the 1st International Workshop on Mobile Location-Based Service (MLBS 2011, in conjunction with Ubicomp). September 17-21, Beijing, 2011: 71-78 .
    • C. Zhou, X. Meng, D. Liu: Spatio-Temporal Sequence Searching in Flickr(Demonstration). In Proceedings of the 34rd Annual International ACM Special Interest Group on Information Retrieval Conference on Research and Development in Information Retrieval (SIGIR2011):1265, July 24-28,2011, Beijing.
    • C. Zhou, X. Meng: STS: Complex Spatio-Temporal Sequence Mining in Flickr. In Proceedings of the 16th International Conference on Database Systems for Advanced Applications(DASFAA2011):208-22, April 22-25, 2011. Hong Kong.
    • C. Zhou, X. Meng: The Researches and Challenges of Complex Event Detection in Pervasive Comptuing. Journal of Frontiers of Computer Science and Technology, Vol. 4, No.12: 1057-1072, Dec. 2010.
    • C. Zhou, X. Meng, J. Wen: Complex Event Detection on Flickr. Journal of Computer Research and Development, Vol.47(Suppl): 1-7, Oct.2010(NDBC2010, Beijing)
    • Y. Huang, X. Pan, X. Meng: OrientPrivacy: An Anonymizer for Privacy Preserving in Mobile Services. Journal of Computer Research and Development, Vol.47(Suppl): 438-441,Oct.2010(NDBC2010, Beijing)
    • C. Zhou, X. Meng: IO3: Interval-based Out-of-Order Event Processing in Pervasive Computing. In Proceedings of the 15th International Conference on Database Systems for Advanced Applications (DASFAA 2010): 261-268, April 1-4, 2010, Japan.
    • X. Pan, X. Hao, X. Meng: Privacy Preserving Towards Continuous Query in Location-Based Services. Journal of Computer Research and Development, Vol.47(1): 121-129, 2010.1.
    • X. Pan, X. Meng, J. Xu: Distortion-based Anonymity for Continuous Query in Location-Based Mobile Services. In Proceedings of the 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL GIS 2009):256-265, November 4-6, 2009, Seattle, Washington. (FULL paper)
    • C. Zhou, X. Meng: Complex Event Detection in Pervasive Computing. The Third SIGMOD PhD Workshop on Innovative Database Research (IDAR2009), June 28, 2009, Providence, USA
    • X. Pan, X. Hao and X. Meng: Privacy Preserving towards Continuous Query in Location-based Services. In Proceedings of the 26th National Database Conference of China(NDBC2009), October 15-18, 2009, Nanchang, China. (in Chinese)
    • Zhiming Ding, Xiaofeng Meng, Shan Wang: O2PC-MT: A Novel Optimistic Two-Phase Commit Protocol for Mobile Transactions. DEXA 2001: 846-856
    • Zhiming Ding, Xiaofeng Meng, Shan Wang: A Transactional Asynchronous Replication Scheme for Mobile Database Systems. J. Comput. Sci. Technol. 17(4): 389-396 (2002)
    • Zhiming Ding, Xiaofeng Meng, Shan Wang: A Novel Conflict Detection and Resolution Strategy Based on TLRSP in Replicated Mobile Database Systems. DASFAA 2001: 234-240
    • Shaoyi Yin, Jidong Chen, Xiaofeng Meng, Caifeng Lai: Storage and Recovery Techniques for PhoneDB. NDBC 2005
    • Xiaofeng Meng, Zhiming Ding: DSTTMOD: A Future Trajectory Based Moving objects Database. DEXA 2003: 444-453
    • Jidong Chen, Xiaofeng Meng, Yanyan Guo, Stephane Grumbach, Haixun Wang: Modeling and Predicting Future Trajectories of Moving objects in a Constrained Network. To appear in MDM workshop-MLASN 2006
    • Jidong Chen, Xiaofeng Meng, Benzhao Li, and Caifeng Lai: Tracking Network-Constrained Moving objects with Group Updates. To appear in WAIM 2006
    • Rui Ding, Xiaofeng Meng, Yun Bai: Efficient Index Update for Moving objects with Future Trajectories. DASFAA 2003: 183-194
    • J. Chen, X. Meng, Y. Guo, X. Zhen: Update-efficient Indexing of Moving Objects in Road Networks. In Proceedings of the Third Workshop on Spatio-Temporal Database Management in conjunction with VLDB 06 (VLDB-STDBM2006), Seoul, Korea, September 11, 2006.
    • J. Chen and X. Meng: Update-efficient Indexing of Moving Objects in Road Networks. accepted for publication in Geoinformatica.
    • X. Hao, L. Wang, X. Meng: Multiple kNN Queries Processing for Moving Objects in Road Networds. Journal of Computer Research and Development, Vol. 44 Suppl.: 113-118, 2007.10 (NDBC2007, Haikou) (in Chinese)
    • X. Hao, X. Meng and J. Xu: Continuous Density Queries for Moving Objects. Proceedings of SIGMOD MobiDE2008, page 1-8, Vancouver, BC, Canada, June 10-12, 2008.
    • C. Lai, L. Wang, J. Chen, X. Meng, K. Zeitouni: Effective Density Queries for Moving Objects in Road Networks. In Proceedings of APWeb/WAIM’07, Huangshan, China, June 17-19, 2007.
    • J. Chen, C. Lai, X. Meng, J. Xu, H. Hu: Clustering Moving Objects in Spatial Networks. In proceedings of the 12th International Conference on Database Systems for Advanced Applications (DASFAA 2007) , Bangkok, Thailand, April 9-12, 2007.
    • X. Meng, S. Yin, Z. Xiao. WDIM: A Web Data Integrated Middleware for LBS. In the Wuhan University Journal of Natural Science (English version) and the third Web Information System and Application. (WISA), November 3-5, 2006.
    • Z. Xiao, X. Meng, J. Xu: Quality Aware Privacy Protection for Location-based Services. In Proceedings of the 12th International Conference on Database Systems for Advanced Applications (DASFAA 2007), page 434-446, Bangkok, Thailand, April 9-12, 2007. ( Full paper: 70/375=18.7%)
    • X. Zhen, J. Xu, X. Meng: p-Sensitivity: A Semantic Privacy-Protection Model for Location-based Services. Proceedings of PALMS2008 associated with MDM2008, Beijing, China, 2008
    • X. Pan, J. Xu, X. Meng: Protecting Location Privacy against Location-Dependent Attack in Mobile Services, accepted by CIKM2008
    • X. Pan, Z. Xiao,X. Meng: Survey of Location Privacy Preserving. Journal of Computer Science and Frontires, Vol.1(3): 268-281, October, 2007
    • J. Chen and X. Meng: Indexing Future Trajectories of Moving Objects in a Constrained Network. Journal of Computer Science and Technology, Vol.22(2): 245--251, March, 2007
    • J. Chen, X. Meng, C. Lai: Clustering Objects in a Road Network. In Journal of Software , Vol.18(2):332-344, February, 2007
    • Y. Bai, Y. Guo, X. Meng, T. Wan, K. Zeitouni : Efficient Dynamic Traffic Navigation with Hierarchical Aggregation Tree. In Proceedings of the The Eighth Asia Pacific Web Conference(APWeb2006), pages 751-758, Harbin, China, January 16-18, 2006. LNCS 3841
    • Zhiming Ding, Xiaofeng Meng, Yun Bai, Rui Ding: Relational Database Supports for location dependent queries. Computer Research and Development, v41, n3, March, 2004, p492-499
    [Top]        
    Maintained by Zhongyuan Wang() Copyright © 2007-2008 WAMDM, All rights reserved