Second International Conference on Computer and Network Technology

Location dependent Query Processing Issues, Challenges and Applications

Kahkashan Tabassum1, Maniza Hijab1

A. Damodaram2

Associate Professors , Dept. of CSE,

Muffakham Jah College of Engineering
Hyderabad 500155, India
{(kahkashan, manizahijab} @mjcollege.ac.in

Professor, Department of CSE,

JNTU College of Engineering,
Hyderabad 500072, India
[email protected]

on their move to some location. This feature has created a

new research area in mobile computing that is contextaware query processing in location-dependent environment.
This research area is concerned with the location-based
query processing applications such as navigation services
for understanding the routes to a destination (using the
digital maps, digital yellow pages directory), emergency
services, location-based advertising for marketing products
of supermarkets, stock exchange information, criminal
tracking, vehicle tracking, animal tracking, lost mobile
tracking, etc. This paper is an attempt to study, understand
and highlight the various issues and challenges that are
encountered during location-based query processing. Since
the result of a LBQ depends on the location of the object the
result of the query keeps on changing very quickly hence it
needs to be adapted very fast to the changes before the
correct result is provided to the concerned user. The result
of the query needs to be continuous and it should match the
movement of the mobile user. Consider the following
queries - a mobile user interested in getting the information
about the location-based data such as the nearest hotel or
taxi. Query Processing taking location of the mobile user
into account was considered in [3] for the first time and
remained as an important topic of research since then till
date. The paper is prepared to cover the issues, data
management challenges of location based query processing
and its applications as follows Section II covers various
aspects of query processing, followed by data management
for carrying out query processing in mobile environment in
section III, section IV highlights the applications of
Location-based query processing application namely
location-based services and finally conclusion is given in
Section V.

Abstract - The growth in the number of mobile users is

motivating the new technology of mobile computing to offer
many new types of services like location-based services that
pose many challenging problems for location-dependent query
processing and data management. In this paper, we discuss the
various challenges in distributed processing of locationdependent continuous queries in a mobile environment by
studying the different scenarios in which both the querying
unit and the object being queried are in motion. Then we
proceed to classify different types of location dependent
queries. We discuss the different solutions to solve these
complex location dependent queries using caching techniques
that facilitate fast access to data and at the same time minimize
the massive network traffic generated due to location
dependent query processing in a wireless environment. To
further reduce the data transfer the query results are cached
so that these results are reusable. The conventional query
processing method like join processing of data from mobile
units and server is directly applied in the mobile environment.
Additionally this paper also focuses on the methods of locationdependent query processing and applications. The main goal of
this paper is to study the various location dependent query
processing methods along with data management problems
involved in evaluation of Location-dependent queries and
classify them.
Keywords: Location dependent query processing (LDQP),
Location-based query (LBQ), Location-based services (LBS),
Mobile User (MU), Invalidation Reports (IR), Value added
Service provider (VASP).

I.

INTRODUCTION

Due to the heavy use of mobile devices both to transmit

voice and data nowadays we find intense research in the
area of mobile communications and networking. The main
aim of the research being facilitating data access to the user
anywhere and at any instant of time. In spite of the
promising advantages and privileges offered by mobile
computing there are a few data management challenges that
are due to the critical limitations pertaining to connectivity
constraints, low battery, unreliable communication, low
memory, slow processing speed, less storage and small
display size. All these factors require that the mobile
devices be adaptable to the latest available environment. An
important feature of mobile computing applications is the
location dependent data access when the mobile devices are
978-0-7695-4042-9/10 $26.00 2010 IEEE
DOI 10.1109/ICCNT.2010.39

II.

LOCATION BASED QUERIES

Location-based queries or location-dependent queries

provide support for Location-based services (LBS).
Location-based applications (LBA) offer location-based
services (LBS) by using queries called the Location-based
Queries (LDQ); the result of execution of these queries is
based on the location of the mobile user (MU). These
queries can be classified into spatial queries or temporal
queries that can be further classified into continuous or non239

from providing great economical benefits the mobile

computing is required to adapt to the existing techniques for
storing data in mobile databases, where the users and data
may be mobile as compared to the conventional distributed
database systems. The mobile environment has a great
impact on the Data management issues namely data
fragmentation, replication among mobile units, consistency
maintenance, etc., Query processing issues namely
approximate answers of location-dependent queries,
optimization of query plans, query fragmentation and
routing, transaction management issues namely relaxation of
the ACID properties due to the frequent disconnections that
may lead to long or blocked transactions. Also user
interfaces and query languages must be adapted to the small
displays of mobile devices.

continuous queries. For example sometimes the MUs may

be interested in getting the information pertaining to the
nearby location such as a hotel or mall (a) within a specific
distance spatial query (non-continuous). (b) With a
specific facility or product - non-spatial and (noncontinuous) or may be interested in the traffic information at
a traffic junction/signal that is present at a distance, all these
queries are based on fixed locations. Sometimes MUs may
be interested in locating an ambulance in an emergency on
the way or hiring a taxi to travel to some other place i.e.
such objects that are in motion. These types of queries are
called adaptive spatial queries. Some of the distinctive types
of LBQs are as follows:
Range Queries (RQ) retrieves the objects located within
a specific region. Range queries can be static range queries
or moving/mobile range queries. If the region is rectangular
then the range queries are known as window queries.
Window queries can be static or dynamic. If the objects of
the query are stationary or moving, they can take any of the
query type (dq, ddt), (dq, ds), (sq, ddt) and (sq, sdt) where (d
dynamic, dt data, s static, q query).
Nearest neighbor (NN) queries [10] are responsible for
getting the objects closest to a specific location. If they are
capable of getting k objects other than the nearest one then
they are called kNN queries. Static database is used to
execute static NN queries and dynamic database is required
to execute dynamic NN queries. A reverse kNN query can
fetch the objects that have a specified location among their k
nearest neighbors. In spatial databases, constrained NN
queries are the queries with a range constraint for the
objects retrieved.
Navigation queries can fetch the mobile user a suitable
path to the destination taking the network traffic into
consideration. [11]. These type of queries use either of the
following forms: a predicate is applied to many objects; e.g.,
nearest neighbor queries or a predicate is checked for every
object with respect to the other one. Some of the Location
dependent queries are also classified on the basis of the
evaluation time and history of movements [12]. These are
called instantaneous queries (snapshot queries), the answer
for which is computed only once and forwarded to the user
at once. Continuous queries are evaluated continuously until
the user decides to terminate them. Persistent queries are
continuous queries that consider both the current state and
the past states of the moving objects.
III.

A. Caching and Prefetching Location-dependent Data

Caching strategies are of extreme help where the data
management in LDQP is concerned. Many popular and
efficient caching strategies emphasize the importance of
future queries in situations where the objects positions
change instantly with the change in their location.
Processing queries on static objects are described in spatial
area (spatial scope) within which the data is known to be
valid [14]. As pointed in Section I, managing data in a
Location-dependent environment is quite challenging. Some
of the challenges being mobile network and communication
constraints, spatial data and user movements. The location
has to be mapped explicitly or implicitly before a user has
the access to the required information. Symbolic Model and
Geometric model [14] differ in their performance levels in
identifying locations. These models represent a location on a
MUs map (a geometric location) into a symbolic model for
querying the objects. Location-dependent queries at this
point can use different indexing and query processing
strategies such as local or non-local and general or simpler.
Local queries depend on the users current location whereas
non-local involves other location. The queries placing
simple conditions based on equating data such as
Download the local traffic in Location-A can be answered
easily. General queries are based on complex conditions.
These can be spatially constrained or non-spatially
constrained and the results of these queries are based on the
valid scopes. The result of such a query is of the form
(query, result). In a mobile environment the user can access
data using on-demand method, broadcasting or by a
combination of both the data caching techniques to improve
performance. In the first method, a mobile user submits a
query and its location to the server [15]. The server returns
the appropriate data individually or by multicast. This type
of data access should be capable of handling four main
issues of Data Placement, Data Replication, Query
Scheduling, Indexing. In Broadcast method data are
broadcast on a public wireless channel and a mobile user
tunes to it as soon as it gets a query. For general queries
broadcast is inefficient because they require complex query-

DATA MANAGEMENT IN LOCATION DEPENDENT QUERY

PROCESSING

Location-dependent data is the data that change

according to the geographical area (location). Section 4
gives different examples of location based services that
access location dependent data. In this section, we present
some challenges posed by mobile computing with respect to
data management issues and discuss the methods used to
solve these issues caching and broadcasting data. Apart

240

periodically [21] or on-demand [23]. Another technique was

proposed by [22] to ascertain that every mobile device
receives the updated value of all data items that have been
changed at the server. The invalidation reports are delivered
periodically to the users and they can be in doze mode
meanwhile. When interested the mobile users retrieve the
data when they become active. This technique does not take
into consideration latency incurred due to size of IR from
the server. Further studies modify this technique[7], the IR
size is optimized such that the latency of
cache
invalidations for a mobile user is reduced and the mobile
devices get the invalidation reports only on demand, [22,
23] uses the method of broadcasting invalidation reports
only on demand and Mobile Service Station (MSS) is used
to communicate with a certain number of MUs on behalf of
the data server and thus MSS reduces the heavy traffic
between MUs and data servers Some applications that
require on demand broadcasting are stock exchange updates,
weather information reports, news delivery and other
applications that are highly dynamic and sensitive. If the
data is not up-to-date it cannot be useful and accessing this
data is undesirable. The solutions addressing the above
illustrated issues have been discussed in [7, 22, 23, and 24].

processing techniques such as join of two relations. Global

Index and Local Index are the two basic broadcasting
strategies.
In the data caching method, a copy of data is kept with
the mobile user. Location-dependent cache invalidation,
cache replacement and prefetching are the three different
issues that are to be considered in this method. Locationdependent cache invalidation can be implemented simple by
attaching the valid scope with the data object. If the data
objects valid scope is far from users current location, its
reusability turns out to be small; therefore farthest data in
replacement is removed [15]. The Replacement techniques
introduced in [15] are Manhattan Distance and Further
Away Replace (FAR). Data is stored beforehand in the
users cache making anticipation for use in the near future
and this method is called Prefetching. For example the
cache manager would pre-fetch shopping information if user
is going to pass a super market [14]. A semantic caching
scheme [17] for nearest neighbor queries uses Voronoi
diagram for caching data. Semantic caching is challenging
because the use of cache to process location dependent
query is complex since only similar queries can reuse the
cached data. The caching supports processing of specific
types queries such as NN queries and range queries but it is
difficult to reuse cache to process complex queries, since
complex join of relations have to be performed in solving
such queries. Thus the cache management is very complex.
Caching by prefetching data [18] beforehand is done so that
the cache reuse is simplified. Data caching for locationbased spatial queries in P2P environments [19] proposes the
solution to process kNN queries. Mobile cooperative
caching [20] depicts a method that is based on sharing the
query results with the database server so that the
communication bandwidth is utilized effectively. The
semantic caching model [13] uses the method of storing the
data related to previously issued queries and using this data
to answer the queries in future also whenever a new query is
received. Due to the caching of this data the incoming
queries can be answered at least partially. Such queries are
later submitted to the server to get the remaining solution.
The semantic caching model can be used in the location
dependent mobile environment to improve the data
availability as well as the performance of data access.

IV.

LOCATION BASED SERVICES

The increasing development of mobile computing has

given rise to a new type of services and applications called
location-based services (LBS). These are also called mobile
location services [1] and are considered as popular
applications of Location dependent query processing. The
extensive developing applications on mobile devices
provides the users with wide range of services. LBS provide
special services by considering the location of the mobile
users and providing them with customized information. The
most popular among them are Value-added services that are
supplied either in-house by the mobile network operators or
by a third-party value-added service provider (VASP), also
known as a content provider (CP). VASPs typically connect
to the operator using protocols like Short message peer-topeer protocol (SMPP), connecting either directly to the short
message service centre (SMSC) or, increasingly, to a
messaging gateway that allows the operator to control and
charge of the content. Average revenue per user abbreviated
to ARPU is a measure used by consumer communications
and networking companies, it is the total revenue divided by
the number of subscribers.This term is used by companies
that offer subscription services to clients for example,
telephone carriers, Internet service providers, hosts and it is
a measure of the revenue generated by one customer phone,
pager, etc. per unit time, typically per year or month. In
mobile telephony, ARPU includes not only the revenues
billed to the customer each month for usage, but also the
revenue generated from incoming calls, payable within the
regulatory interconnection regime. Among mobile
computing applications provided to the users in the form of
value added services, location-based services are expected

B. Broadcasting Location-dependent Data

The most economic and effective technique of
disseminating data to a large number of users on mobile
communication network is broadcasting. The methods
depicted in [22, 23] use a push-based data delivery
technique of data distribution in which the server broadcasts
invalidation report in which the modified data items are
indicated. The mobile users directly listen to the broadcast
channel to get the validated cached copies instead of
querying the server to validate their local cache. The server
can broadcast the Invalidation Reports (IR) either

241

to be a big business for wireless operators. This paper

summarizes a method to provide these services considering
the location of the user and also gives a method to the user
that provides him with customized information. This can be
achieved by the processing of continuous locationdependent queries by establishing an uplink and downlink
connection with the server that provides required service
instead of providing through SMS. The service can be
provided on demand from the user after the user queries the
server. One of the key issues is how to update the answer
presented to the user efficiently as it becomes vulnerable
and outdated with time very quickly. Although uneconomic
LBS provides other advantages such as more entertainment
and productivity. The demand of LBS is increasing
everyday.
The figure below is a modified LBS architecture
[8] that is used to serve the mobile users requesting both
location-dependent and location-independent data. The
location-based architecture in the figure below explains the
services provided from the point when the mobile user
initiates the request to a particular location based service to
the response that is broadcasted to the user in the form of
response. The LBS system is composed of three major units
- mobile user, central database server, local database server
that comprises of components location services & profile
manager, query generator, query processor, query optimizer,
remote data manager, local cache manager, query cache
manager and a data integrator that performs the function of
keeping the overall data in place together.

There is a local database server for every cell that acts as an

interface between the central database server and mobile
user. The central database server is used to create the service
site for data collected. A push unit is used to store
prefetched data as required by the server. This process
reduces cost of uplink communication thus utilizing the
wireless channel effectively. Some examples of real time
location-based services are given below:
Emergency services: Some LBS provide roadside
assistance by tracking the GPS location of a vehicle and
provides assistance in locating the individual who is
unaware of his/her location in an emergency situation. A
technique called geographic-based messaging [2] is used to
warn about danger to the users within a certain area
Navigation and information services are services like
digital travel assistants that provide driving directions based
on time taken to travel, distance covered. Some other
examples are location-dependent web pages [3] and mobile
yellow pages that provide information about nearby points
of interest [4].
Advertising applications:
In mobile commerce, the
concept of location dependent advertising, or proximitytriggered advertisements, is an interesting and efficient way
of inviting users to shopping. [5]. Many people can be
attracted by advertising e-discounts that motivate them to
perform purchases inside the shopping malls.
Tracking services provides assistance to the users in
tracking the present locations of moving objects like
vehicles, friends, animals in a sanctuary, for parents to track
their children, old and retarded people, products, or
personnel, tracking suspected criminals and specific
employees, pets, monitoring taxis in an area ). The
flexibility in use of GPS devices plays an important role in
the development of these services. The GPS devices can be
worn on clothing or can be fixed in bracelets or watches.
Billing services help the service providers to allow
individual subscribers to perform billing using locationbased services with their convenience as per the cost
sensitivity of the location. Thus a new method of billing
procedure is promoted by this type of application where the
products are bought without paying directly to the cashier
but the payment done automatically.
In Location-based games the players move in the real
world and interact with each other and with the environment
depending on their locations. Some of them are as follows
racing in space, Swordfish, etc. These games depend on
positioning systems. Some of the other types of location
based services are location-based message boards through
which user can post messages within there vicinity.
V.

CONCLUSION

The in exhaustive and intensive development in the area

of mobile computing, location prediction and wireless
communications has given rise to information services that
are able to predict, identify and adapt to the location of the
mobile users. The location dependent query processing is

242

International Database
(IDEAS03).

the building block for the location based services. Locationbased services consist of potential applications that span
every context in daily life like responses to natural and
human-induced hazards in a definite and proper manner.
These services are helpful in many ways such as they
provide mobile users with the information rapidly and
accurately about the immediate surroundings that are
required by a specific user in a specific context. Yet the
research challenges in this are incomplete and in exhaustive.
Few challenging applications in progress are devising
infrastructure for location-sensing, adaptive prediction
methods and mechanisms for mobile clients where location
dependent processing is concerned; extending maximum
support for privacy protection; and various adaptive
technologies for representing, monitoring and displaying
location dependent information any time and anywhere even
if the communication and display facilities are highly
scarce. The future of these location dependent services is
highly sensitive for rapid development that requires
promising investments, but at the same time the future of
well-being of these services is significantly critical.

Engineering

and

Applications

Symposium

[9] Seydim, A. Y., Dunham, M. H., and Kumar, V., Processing locationdependent queries in a multi-cell wireless environment, 2nd ACM
International Workshop on Data Engineering for Wireless and Mobile
Access (MobiDE01), ACM, USA, 4753.
[10] Tao Y. and Papadias D., Efficient historical R-trees, 13th IEEE
International Conf. on Scientific and Statistical Database Management
(SSDBM01), USA, 223232.
[11] Lee W. and Zheng B., A fully distributed spatial index for locationbased wireless broadcast services, 25th IEEE nternational Conf. on
Distributed Computing Systems (ICDCS05), USA, 349358.
[12] Sistla A. P., Wolfson, O., Chamberlain S., and Dao S., Modeling and
querying moving objects, 13th IEEE International Conf. on Data
[13] Dar S., Franklin M. J., Jonsson B. T., Srivastava D., and Tan M.,
Semantic data caching and replacement, 22nd International Conf. on Very
Large Data Bases (VLDB96) USA, 330341.
[14] Jianliang Xu, Xueyan Tang and Dik L. Lee, Performance Analysis of
location dependent cache invalidation schemes for Mobile Environments,
IEEE Transactions on Knowledge and Data Engineering, April 2003.
[15] Baihua Zheng, Jianliang Xu, Dik L. Lee, Cache Invalidation and
replacement strategies for Location-dependent data in Mobile
Environments, IEEE Transactions on Computers, Oct 2002.

ACKNOWLEDGMENT
We would like to extend our gratitude to Dr. Khaleelur
Rahman, Head Dept. of Computer Science, MJCET for
giving his guidance in carrying out this research work.
Thanks are also due to Mr. Mujahid Ahmed Ali for
motivating us to pursue this research.

[16] Zheng B., Lee W., and Lee D. L., Spatial index on air, 1st IEEE
International Conf. on Pervasive Computing and Communications
(PerCom03), USA, 297304.
[17] Zheng B. and Lee D. L., Semantic caching in location-dependent
query processing, 7th International Symposium on Advances in Spatial and
Temporal Databases (SSTD01), Springer, Berlin lberg, 97116.

REFERENCES

[18] Hu H., Xu J., Wong W. S., Zheng B., Lee D. L., and Lee W.C.,
Proactive caching for spatial queries in mobile environments, 21st IEEE
International Conf. on data Engg. (ICDE05), USA, 403414.

[1] Giaglis, G. M., Kourouthanassis, P., Tsamakos, A., Towards a

classification framework for mobile location services, Mobile commerce:
technology, theory, and applications, IDEA-04 Group Publishing,
Hershey, PA, 6785.

[19] Ku W., Zimmermann R., and Wan C., Location-based spatial queries
with data sharing in mobile environments, Exploiting location information
for info station-based Hoarding, 7th International Conf. on Mobile
Computing and Networking (MobiCom01). ACM, USA, 1527.

[2] Basagni, S., Chlamtac, I., and Syrotiuk, V. R., Geographic messaging in
wireless ad hoc networks, 49th IEEE Vehicular Technology Conf.
(VTC99). USA, 19571961, 2003.

[20] Chow C., Leong H. V., and Chan A. T. S, Peer-to-peer cooperative

caching in a hybrid data delivery environment, 7th IEEE International
Symposium on Parallel Architectures, Algorithms, and Networks
(ISPAN04), USA, 7985.

[3] Haghighat, A., Lopes, C., Givargis, T., Mandal, A., Location-aware
web system, OOPSLA04.
[4] Chon, H. D., Agrawal, D., and Abbadi, NAPA:Nearest available
parking lot application,
18th IEEE International Conf. on Data
Engineering (ICDE02), USA, 496497.

[21] Prof. M. Satyanarayanan, Consistent Data Availability (CODA),

Distributed Computing, 2001.

[5] Alonso, R. and Korth, H. F., Database system issues in nomadic

computing, International Conf. on Management of Data (SIGMOD93)
ACM, USA, 388392.

[22] Barbara, Imielinski, Sleepers and Workaholics: Caching Strategies in

Mobile Environments, Proc. Of ACM, SIGMOD, May 1995.

[6] Imielinski, T. and Nath, B., Data management for mobile computing,
SIGMOD, (Mar 93), 3439.

[23] Prof. Daniel Barbara , Asynchronous Stateful, Distributed Computing,

2001.

[7] Jing, Elmagarmid, Bit Sequences: An Adaptive Cache Invalidation

method in client / server environment, ACM (MONET) vol. 2.

[24] Zhijun Wang, Sajal Das, Hao Che and Mohan Kumar, A Scalable
Asynchronous Cache Consistency Scheme (SACCS) for Mobile
Environments, IEEE Transactions on Parallel and Distributed Systems,
Vol. 15 Pg: 983 995, November 2004.

[8] Shiow-yang, Wu Kun-Ta Wu, Dynamic Data Management for Location

Based Services in Mobile Environments, Proceedings of the Seventh

243