Crowdsourced Trace Similaritywith Smartphones

ABSTRACT:

Smartphones are nowadays equipped with a number of sensors, such as WiFi, GPS, accelerometers, etc. This capabilityallows smartphone users to easily engage in crowdsourced computing services, which contribute to the solution of complex problemsin a distributed manner. In this work, we leverage such a computing paradigm to solve efficiently the following problem: comparing aquery trace Q against a crowd of traces generated and stored on distributed smartphones. Our proposed framework, coinedSmartTraceþ, provides an effective solution without disclosing any part of the crowd traces to the query processor. SmartTraceþ, relieson an in-situ data storage model and intelligent top-K query processing algorithms that exploit distributed trajectory similaritymeasures, resilient to spatial and temporal noise, in order to derive the most relevant answers to Q. We evaluate our algorithms onboth synthetic and real workloads. We describe our prototype system developed on the Android OS. The solution is deployed over ourown SmartLab testbed of 25 smartphones. Our study reveals that computations over SmartTraceþ result in substantial energyconservation; in addition, results can be computed faster than competitive approaches.

EXISTING SYSTEM:

In our previous work, we have already paved theway toward trajectory processing techniques in a distributedmanner (i.e., without percolating each and every usergeolocation to a central authority.) However, those wereboth agnostic in terms of energy and time constraints thatarise in a smartphone network, but also in respect to thetrajectory trace disclosure issues (i.e., they assumed that thequery processor can arbitrarily access the distributedtrajectories.)

DISADVANTAGES OF EXISTING SYSTEM:

Services assume that the user trajectoriesare stored on a centralized or cloud-like infrastructure priorto query execution.

PROPOSED SYSTEM:

In this paper, we present a crowdsourced tracesimilarity search framework, called SmartTraceþ, whichenables the execution of queries in the form: “Report theusers that move more similar to Q, where Q is some querytrace.” The notion of similarity captures the traces (i.e.,trajectories) that differ only slightly, in the whole sequence,from the query Q. Our framework enables the execution ofsuch queries in both outdoor environments (using GPS)and indoor environments (using WiFi Received-Signal-Strength), without disclosing the traces of participatingusers to the querying node

ADVANTAGES OF PROPOSED SYSTEM:

1. Smartphones have both expensive communicationmediums but also asymmetric upload/downloadlinks, thus by continuously transferring data to thequery processor can both deplete the precioussmartphone battery faster, increase user-perceiveddelays, but can also quickly degrade the network health

2. Continuously disclosing user positional data to acentral entity might compromise user privacy inserious ways. This creates services that haverecently raised many concerns, especially for socialnetworking services (e.g., Facebook, Buzz, etc.)and smartphone services

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

System: Pentium IV 2.4 GHz.

Hard Disk : 40 GB.

Floppy Drive: 1.44 Mb.

Monitor: 15 VGA Colour.

Mouse: Logitech.

Ram: 512 Mb.

MOBILE:ANDROID

SOFTWARE REQUIREMENTS:

Operating system : Windows XP.

Coding Language: Java 1.7

Tool Kit:Android 2.3

IDE:Eclipse

REFERENCE:

Demetrios Zeinalipour-Yazti, Member, IEEE, Christos Laoudias, Student Member, IEEE,Constandinos Costa, Michail Vlachos, Maria I. Andreou, and Dimitrios Gunopulos, Member, IEEE, “Crowdsourced Trace Similaritywith Smartphones”, IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING, VOL. 25, NO. 6, JUNE 2013