Uploaded on


  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Be the first to comment
    Be the first to like this
No Downloads


Total Views
On Slideshare
From Embeds
Number of Embeds



Embeds 0

No embeds

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

    No notes for slide


  • 1. J.Kartheeswari / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue 4, July-August 2012, pp.385-389 Implementation Of Location Monitoring Services Based On Anonymization Algorithm J.Kartheeswari Centre For Information Technology And Engineering, M S University, TirunelveliABSTRACT Anonymization algorithm is mainly sensors and counting sensors. In identity sensorused to monitor the location . In this paper location monitoring system, the sensor nodes reportpropose an implementation of location the exact location information of the monitoredmonitoring services based on resource and persons to the server. While counting sensorquality aware algorithm. In resource aware monitoring system, each sensor node reports thealgorithm to minimize time and communication number of objects in its sensing area to the server.cost. In existing system to find the minimum We propose two anonymization algorithm namelybounding rectangle using monitor area. In our Resource-aware and Quality-aware algorithm. Inpaper another way to find the minimum Resource aware algorithm to minimizebounding rectangle using monitor object. While communication cost. In quality aware algorithm tofinding the minimum founding rectangle with provide accurate location.monitor object and monitor area we find thatthe time to process is equal. II. SYSTEM ARCHITECTUREKeywords – Wireless sensor networks, Location Server adminprivacy, Aggregate location AggregateI. INTRODUCTION Set anonymity location info In wireless sensor network (WSN) is an valuead-hoc network composed of small sensor nodesdeployed in large numbers to sense the physical Sensor Node Sensor Nodeworld. Wireless sensor networks have very broadapplication prospects including both military andcivilian usage. In mobile sensor network Query Locationdevelopment of algorithms and prototype vehicles updatefor wide-area surveillance and reconnaissanceusing mobile sensor networks (MWSN). User User User UserMonitoring on land, water and air using largenumbers of mobile sensor nodes is demonstrated atour Distributed Intelligence and Autonomy Lab Sensor Nodes: Each sensor node is responsible for(DIAL). Mobile sensor networks are sensor determining the number of objects in its sensingnetworks in which nodes can move under their own area. Sensor nodes blurs its sensing area into acontrol or under the control of the environment. cloaked area, which includes at least k objects, andMobile networked systems combine the most reports with the number of objects located inadvanced concepts in perception, communication, particular region as an aggregate locationand control to create computational systems information to the server. Each sensor node is alsocapable of interacting in meaningful ways with the aware of its location and sensing area.physical environment, thus extending the individualcapabilities of each network component and Server: Server collects the aggregate locationsnetwork user to encompass a much wider area and reported from the sensor nodes, using a spatialrange of data. A key difference between a mobile histogram to estimate the distribution of thesensor network and a static sensor network is how monitored objects. Also server answers rangeinformation is distributed over the network. Under queries raised by users, based on the estimatedstatic nodes, a new task or data can be flooded object distribution. Administrator can change theacross the network in a very predictable way. anonymized level k of the system at anytime byUnder mobility this kind of flooding is more disseminating a message with a new value of k tocomplex. Under natural mobility this depends on all the sensor nodes.the mobility model of the nodes in the system. Forthe location monitoring system using identity Users: Each and every user updates their location information to the sensor node. Users can issue 385 | P a g e
  • 2. J.Kartheeswari / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue 4, July-August 2012, pp.385-389range queries to the system through the sensor of objects located in its sensing area m.count, to itsnodes. They can get reply for query like, what is neighbors. When m receives a message from a peerthe number of persons in a certain area? The server p, m stores the message in its PeerList. Wheneveruses the spatial histogram to answer their queries. m finds an adequate number of objects, m sends a notification message to its neighbors. If m has notPrivacy model: Sensor nodes constitute a trusted received the notification message, some neighborszone, communicate with each other through a has not found an adequate number of objects,secure network channel to avoid internal network therefore m forwards the received message to itsattacks, for example, eavesdropping, traffic neighbors.analysis, and malicious nodes. The system providesanonymous communication between the sensor 3.2 Cloaked Area Stepnodes and the server by employing existing Cloaked area step is that each sensor nodeanonymous communication techniques. Thus given blurs its sensing area into a cloaked area thatan aggregate location R, the server only knows that includes alteast k objects to satisfy the k-anonymitythe sender of R is one of the sensor nodes within R. privacy requirement. To minimize computationalAuthenticated administrators can change the k- cost, this step uses a greedy approach to find aanonymity level. Administrators can set the k- cloaked area based on the information stored inanonymity level to a small value to get more PeerList.accurate aggregate locations from the sensor nodes, 3.2.1 Scoreor even set it to zero to disable the algorithm to get The score is defined as a ratio of thethe original readings from the sensor nodes, in object count of the peer to the euclidean distanceorder to get the best services from the system. This between the peer and m. The idea behind the scoreis a nice privacy-preserving feature, because the is to select a set of peers from PeerList to S to formobject count of a small area is more likely to reveal a cloaked area that includes at least k objects andpersonal location information. The definition of a has an area as small as possible. Then, wesmall area is relative to the required anonymity repeatedly select the peer with the highest scorelevel, because our system provides lower quality from the PeerList to S until S contains at least kservices for the same area if the anonymized level objectsgets stricter.Aggregate Location: Each sensor node blurs itssensing area into a cloaked area, in which at least kpersons are residing. Each sensor node reports onlyaggregate location information, which is in a formof a cloaked area A, along with the number ofpersons, N, located in A, where N ≥ k, to theserver. A smaller k indicates less privacyprotection, because a smaller cloaked area will bereported from the sensor node; hence bettermonitoring services. A larger k results in a largercloaked area, which will reduce the quality ofmonitoring services, but it provides better privacyprotection.III. RESOURCE AWARE ALGORITHM Figure 2. Resource aware cloaked area of Resource aware algorithm indicates that sensor Athe sensor nodes can communicate directly witheach other. This algorithm consists of three steps. Figure 2. illustrates the cloaked area step. The PeerList of sensor node A contains the3.1 Broadcast Step information of three peers, B, D, and E. The object Broadcast step is to guarantee that each count of sensor nodes B, D, and E is 3, 1, and 2,sensor node knows an adequate number of objects respectively. We assume that the distance fromto compute a cloaked area. To reduce sensor node A to sensor nodes B, D, and E is 17,communication cost, this step relies on a heuristic 18, and 16, respectively. The score of B, D, and Ethat a sensor node only forwards its received is 3/17 = 0:18,1/18 =0:06, and 2/16 = 0:13,messages to its neighbors when some of them have respectively. Since B has the highest score, wenot yet found an adequate number of objects. select B. The sum of the object counts of A and B In this step, after each sensor node m is six which is larger than the required anonymityinitializes an empty list PeerList, m sends a with its level k = 5, so we return the MBR of the sensingidentity m.ID, sensing area m.Area, and the number area of the sensor nodes in S, i.e., A and B, as the 386 | P a g e
  • 3. J.Kartheeswari / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue 4, July-August 2012, pp.385-389resource-aware cloaked area of A, which is rectangle. In existing system to find the minimumrepresented by a dotted rectangle. bounding rectangle using monitor area. In our3.2.2 Minimum Bounding Rectangle paper another way to find the minimum bounding For each sensor node initializes in its rectangle using monitor object.PeerList. It includes atleast k-objects and has anarea as small as possible. Finally, m determines thecloaked area that is a minimum boundingrectangle(MBR) that covers the sensing area of thenodes, and the total number of objects. An MBR isa rectangle with the minimum area that completelycontains all desired regions.3.3 Validation Step In validation step is to avoid reportingaggregate locations with a relationship to serverEach sensor node maintains a list to store theaggregate locations sent by other peers. AS thisstep ensures that no aggregate location with thecontainment relationship is reported to the server, Figure 3. Quality aware cloaked area ofthe adversary cannot obtain any deterministic sensor Ainformation from the aggregate locations. Since the Figure 3 illustrate the area of MBRserver receives an aggregate location from eachsensor node for every reporting period, it cannot A, E is less than current minimal cloaked areatell whether any containment relationship takes and the total number of monitored objects in MBRplace among the actual aggregate locations of the A, E is k= 5, we set A, Eto the currentsensor nodes. minimal cloaked area 4.2.1 Monotonicity propertyIV QUALITY AWARE ALGORITHM This property propose two pruning The Quality-aware algorithm initializes a conditions in the lattice structure. 1. If thevariable current minimal cloaked area. When the combination gives the current minimal cloakedalgorithm terminates, the current minimal cloaked area, other combinations that contains at the higherarea contains the set of sensor nodes. This levels of the lattice structureshould be pruned. 2. Ifalgorithm consists of three steps. a combination constitutes a cloaked area that is the same or larger than the current minimal cloaked4.1 Search Space Step area, other combinations that contain at the higher The search space step is too costly for levels of the lattice structure should be pruned.node m to gather the information of all the sensornodes to compute its minimal cloaked area. To 4.3 Validation Stepreduce communication and computational cost, m This step is exactly the same as in thedetermines a search space based on the input initial resource-aware algorithm.solution. It is to compute the minimal cloaked area. V IMPLEMENTATION4.2 Minimal Cloaked Area Step In this application, we try to implement Minimal cloaked area takes a set of peers location monitoring system. We use MS access asin search space, computes the minimal cloaked area database for this application. We are going tofor the sensor node. It propose two optimization develop a location monitoring system, where usertechniques to reduce computational cost. The first updates their location to server through sensoroptimization technique is that need not to examine node. Sensor node cloaks the exact location ofall the combinations of the peers. This optimization client to region coverage range, thus the privacy ofmainly reduces computational cost by reducing the the user can be preserved. Also more privacy cannumber of computations among the peers. The be achieved by using k-anonymity value, which cansecond optimization technique has two properties be set by admin. More the value of k-anonymity 1. Lattice Structure means more privacy for users. The users in 2. Monotonicity Property particular region can rise query to server about the4.2.1 Lattice structure: number of users in that particular region. In our Lattice structure is used to generate the system, the sensor nodes constitute a trusted zone,combinations of the sensor nodes. It generates the and communicate with each other through a securelattice structure from the lowest level based on a network channel to avoid internal network attacks,simple generation rule. In lattice structure concept for example, eavesdropping, traffic analysis, andused for to finding the minimum bounding malicious nodes. 387 | P a g e
  • 4. J.Kartheeswari / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue 4, July-August 2012, pp.385-389 The aggregate location using k-anonymity REFERENCESvalue can arrived in the coming phase. Also the [1] D. Culler and M.S. Deborah Estrin,coming phase work includes, the given aggregate “Overview of Sensor Networks,” Computer,location R, the server only knows that the sender of vol. 37, no. 8, pp. 41-49, Aug. 2004.R is one of the sensor nodes within R. Furthermore,only authenticated administrators can change the k- [2] Kido, H.;Yanagisawa,Y.; Satoh, T. “ Ananonymity level and the spatial histogram size. In anonymous communication technique usingemergency cases, the administrators can set the k- dummies for location-based servicesanonymity level to a small value to get more Pervasive Services” ICPS 05. Proceedings.accurate aggregate locations from the sensor nodes. International Conference on Publication ,pp:Since the server and the system user are outside the 88 – 97,2005trusted zone, they are untrusted. We now discussthe privacy threat in existing location monitoring [3] P. Kalnis, G. Ghinita, K. Mouratidis, and In an identity-sensor location monitoring Papadias, “Preventing Location-Basedsystem, since each sensor node reports the exact Identity Inference in Anonymous Spatiallocation information of each monitored object to Queries,” IEEE Trans. Knowledge and Datathe server, the adversary can pinpoint each object’s Eng., vol. 19, no. 12, pp. 1719-1733, Dec.exact location. On the other hand, in a counting- 2007.sensor location monitoring system, each sensornode reports the number of objects in its sensing [4] C.-Y. Chow, M.F. Mokbel, and X. Liu, “Aarea to the server. The adversary can map the Peer-to-Peer Spatial Cloaking Algorithm formonitored areas of the sensor nodes to the system Anonymous Location-Based Services,”layout. If the object count of a monitored area is Proc. 14th Ann. ACM Int’l Symp. Advancesvery small or equal to one, the adversary can infer in Geographic Information Systems (GIS),the identity of the monitored objects based on the 2006.mapped monitored area. [5] Jiejun Kong , Xiaoyan Hong, ANODR: We Well established k-anonymity privacy, anonymous on demand routing withthat is, a person is indistinguishable among k untraceable routes for mobile ad-hocpersons. Enables trusted sensor nodes and provides networks,2005the aggregate location information of monitoredpersons .Each aggregate location is in a form of a [6] C. Bettini, S. Mascetti, X.S. Wang, and S.monitored area A along with the number of Jajodia, “Anonymity in Location-Basedmonitored persons residing in A, where A contains Services: Towards a General Framework,”at least k persons. The resource-aware algorithm Proc. Int’l Conf. Mobile Data Managementaims to minimize communication and (MDM), 2007.computational cost .Quality-aware algorithm aimsto maximize the accuracy of the aggregate [7] B. Gedik and L. Liu, “Protecting Locationlocations by minimizing their monitored areas. Privacy with Personalized K-Anonymity:While finding the minimum bounding rectangle Architecture and Algorithms,” IEEE Trans.with monitor object and monitor area we find that Mobile Computing, vol. 7, no. 1, pp. 1-18,the time to process is equal. Jan. 2008.CONCLUSION [8] T. Xu and Y. Cai, “Exploring Historical In this paper we propose implementation Location Data for Anonymity Preservationof location monitoring services based on in Location-Based Services,” Proc. IEEEanonymization algorithm. In our system, sensor INFOCOM, 2008.nodes execute our location anonymizationalgorithms to provide k-anonymous aggregate [9] G. Ghinita, P. Kalnis, A. Khoshgozaran, C.locations, in which each aggregate location is a Shahabi, and K.-L. Tan, “Private Queries incloaked area A with the number of monitored Location Based Services: Anonymizers Areobjects, N, located in A, where N  k, for the Not Necessary,” Proc. ACM SIGMOD,system. The resource-aware algorithm aims to 2008.minimize communication and computational cost,while the quality-aware algorithm aims to minimize [10] S. Guo, T. He, M.F. Mokbel, J.A. Stankovic,the size of cloaked areas in order to generate more and T.F. Abdelzaher, “On Accurate andaccurate aggregate locations. While finding the Efficient Statistical Counting in Sensor-minimum bounding rectangle with monitor object Based Surveillance Systems,” Proc. Fifthand monitor area we find that time to process is IEEE Int’l Conf. Mobile Ad Hoc and Sensorequal. Systems (MASS), 2008. 388 | P a g e
  • 5. J.Kartheeswari / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue 4, July-August 2012, pp.385-389[11] K. Bohrer, S. Levy, X. Liu, and E. Schonberg, “Individualized Privacy Policy Based Access Control,” Proc. Sixth Int’l Conf. Electronic Commerce Research (ICECR), 2003.[12] E. Snekkenes, “Concepts for Personal Location Privacy Policies,” Proc. Third ACM Conf. Electronic Commerce (EC), 2001.[13] L. Sweeney, “Achieving k-Anonymity Privacy Protection Using Generalization and Suppression,” Int’l J. Uncertainty, Fuzziness and Knowledge-Based Systems, vol. 10, no. 5, pp. 571-588, 2002.[14] H. Kido, Y. Yanagisawa, and T. Satoh, “An Anonymous Communication Technique Using Dummies for Location-Based Services,” Proc. Int’l Conf. Pervasive Services (ICPS), 2005.[15] B. Bamba, L. Liu, P. Pesti, and T. Wang, “Supporting Anonymous Location Queries in Mobile Environments with Privacygrid,” Proc. Int’l Conf. World Wide Web (WWW), 2008. 389 | P a g e