2010 International Conference on Multimedia Information Networking and Security

Geospatial Information Security Web Service based on Agent

†
Jing Sun, Jin Liu, Zhengquan Xu , Zhe Wu
State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing,
Wuhan University, Hubei Wuhan 430079 China
†
e-mail: [email protected], [email protected], [email protected]

Abstract—With the development of remote sensing and architecture. Lee [2] developed a solution for a specific
technology, the use of remote sensing images collected by space platform-service-oriented architecture and web services and
platforms is becoming more and more pervasive. There is more some system modeling strategy. Hwang et al. [3] researched
and more need to distribute processing of remote sensing the dynamic web service selection problem in a failure-
images. At present, WebService can figure out non-security prone environment and proposed two strategies to select
remote sensing images network processing, but due to network web services that are likely to successfully complete the
transmit without non-encryption at client side, which may lead execution of a given sequence of operations. The literatures
to image information leakage. To settle this problem, we do not consider the security of remote sensing images web
propose a geospatial information security Web service based processing.
on mobile agent and WebService. This method can guarantee Distributed processing service of remote sensing images
security image security network processing, which is useful provides good solution to share and processing of non-
complent of WebService. security image, but with regard to the security of image
there is no guarantee due to network transmission without
Keywords-geospatial information security; WebService;
any protection measures. To solve the problems, in this
mobile agent
paper, based on the perspective of the characteristics of
I. INTRODUCTION remote sensing images and image network security
transmission, we propose mobile agent based geospatial
With the rapid development of remote sensing information security service, making that the distributed
technology, the different departments through a variety processing are suitable for any remote sensing images
ways acquire amount of remote sensing image data in high- including security images and the security of image through
speed growth. In addition, due to the various types of web processing can be guaranteed.
remote sensing images and the improvement of The rest of the paper is arranged as follows. In Section
corresponding image resolution, the traditional image II, the distributed processing of remote sensing images, web
processing methods are not available for current application. security transmission of image and mobile agent technology
At present, more pervasive application of remote sensing are introduced. This is followed, in Section III, the web
images and the rapid development of network technology, security service based on mobile agent and WebService is
the demand of the release and processing services of remote presented in detail. At the same time, the experiment is
sensing images are urgent. People are eager to in a wider drawn. Finally, the conclusions and future work are given.
range of share and processing of remote sensing image
through the network. II. RLEATED WORKS
Relative to the general image, remote sensing images
have a large volume of data, the complex processing, and so A. Distributed processing of remote sensing images
on. From the algorithm point of view, the number of remote For distributed processing of remote sensing images, the
sensing images processing algorithms are in large and many C/S model and the B/S model are more pervasive in early
algorithms have targeted, not universal, such as some stage. That is, one or several computers are as server which
extraction algorithm is effective only on the extraction of can provide the users required services, and the client side
building, but extraction of metal is poor, and vice verse. The by installing software or the browser can access these
goal to find one algorithm to effectively extract all types of services. This model, which is simple and easy
ground objects is almost impossible [1]. As any a computer implementation, has played an important role in distributed
and a remote sensing images processing software can not set processing network of GIS [1]. But there are different
all the image processing algorithms in together, so for the between remote sensing images processing and GIS data
remote sensing images processing, in particular, the large processing in general. The former often involves a great
amount of network computing for high-resolution remote amount of data and computation, the server takes much
sensing images processing, distributed environment model longer CPU time and large-scale data throughput;
need to be constructed. The WebService model is put particularly if the number of the visited users is more and
forward as a distributed network computing way and also be multiple tasks are submitted simultaneously to the server,
applied to more and more areas. Shen et al. [1] analyzed the the server often does not normally work because of the big
resource sharing model of grid computing and web service, amount of data and calculation, and network congestion,
and then proposed remote sensing images processing model which may be figured out by WebService model.

978-0-7695-4258-4/10 $26.00 © 2010 IEEE 288

283
DOI 10.1109/MINES.2010.69
Authorized licensed use limited to: JADAVPUR UNIVERSITY. Downloaded on August 14,2023 at 10:15:27 UTC from IEEE Xplore. Restrictions apply.
 WebService technology is a standard mechanism
through the web distribution and use of software services. Content
WebService can easily realize the data exchange and
Leakage
integration between different systems. WebService model
provides, in a scalable, loosely coupled and non-specific
platform environment, the ability to exchange information
by using some standard protocol, such as HTTP, XML,
XSD, SOAP, WSDL, etc.. The main advantage of
WebService is to achieve interoperability of heterogeneous
platforms, which is very necessary to remote sensing images
processing. The WebService distributed system architecture Figure 2. Processing image content leakage.
is as shown in Fig. 1. In WebService model, UDDI is as the
release medium for the provision of services. Different In addition, traditional images encryption transforms
services can have different service provider, that is, different image content into an unintelligible form under the control of
tasks can be accomplished by different computers. It is the key like file encryption. If we encrypt the processing
helpful to reduce the burden of single mode server and remote sensing images with a traditional cipher directly and
enhance the ability of large amount of data processing and completely. This kind of method has high computing
multi-task submitted to support parallel processing in complexity since it encrypts the whole image data stream [4].
remote sensing images processing. In order to improve encryption efficiency under the certain
security condition, we may adopt the encryption method of
content-based remote sensing images during the encryption
process of remote sensing images. Content-based method of
encryption maintains the syntax structure of the ciphertext
unchanged with the plaintext, only some of the semantic
information which plays an important role in understanding
image is encrypted to realize the content-based remote
sensing images protection [5].
C. Mobile agent
The application of agent technologies has been more and
more widely used with the development of network and
artificial intelligence at present, such as artificial
intelligence, databases, and computer networks [6]. The
Figure 1. The WebService architecture.
agent is the simulation of human behaviors and social
relationships, and it could independently run and provide the
WebService follows the design concept of "Register - appropriate services intelligently. Compared with the
find - bind": popular software entity, such as object, component, agent
1) first of all, data services are registered in the has much higher intelligent degree and easy been extended.
directory centers(UDDI); Mobile agent technology is the one of the Agent
2) and then the client log on to the directory center to researcher fields which has been provided a new method of
find services, as shown a line in Fig.1; distributed computing technology. The most striking feature
3) finally, the client bind with the service node for of mobile agent is mobility. In addition, it also inherited the
remote sensing data, as shown b line in Fig.1. property of agent, such as autonomy, reactivity,
asynchronism, intelligence and other key features [7].
B. Security transmission of image
To the existing WebService systems, there exists a big III. SECURITY SERVICE BASED ON AGENT AND
threat, i.e., content leakage. As shown in Fig. 2, the web WEBSERVICE
processing of remote sensing images may be leaked out In order to settle the possible security leakage problem
from the gap between the client side and service provided of web processing, we combine mobile agent with
side. Thus, the image may be obtained by hacker including WebService to compose a security and integrated web
the hostile. A possible solution is to encrypt remote sensing service workflow, which can smartly guarantee the security
images needing to be web processed, which may be based of to-be processed information data from end to end in the
on image format on the condition of the requirement of web.
certain security. However, since encryption must be
achieved at the client side for network security transmission, A. Mobile agent security service architecture
there is need to client encryption software, which increases We can think of security service framework based on
the client burden. MAS (Mobile Agent System) in the web as an agent
community, which is constituted by a variety of self-service
and communication and cooperation with each agent

284
289

Authorized licensed use limited to: JADAVPUR UNIVERSITY. Downloaded on August 14,2023 at 10:15:27 UTC from IEEE Xplore. Restrictions apply.
individuals. Agent security service mainly includes two barriers caused by heterogeneous network, and
parts, i.e., encryption agent service and decryption agent ensuring data security and integrity.
service. Each part can be made up of several
encryption/decryption types, such as parameters selection B. Workflow of security service
and strength degree of encryption/decryption. The In the proposed scheme, when the client side want to
architecture is as shown in Fig. 3. protect web processing remote sensing images to be
processing by web service. It should follow the steps as
follows. Firstly, the client can choose different encryption
parameters through UIA, then, encrypt image with IEA.
During this time, UIA calls the corresponding image
processing service and transmit encryption image data after
channel encode. The server side calls the corresponding
image decryption service. And then, the server processes the
decryption image to meet the client side requirement. When
processing is finished, the server encrypts post-processing
image and transmit the image to the client side. The client
side finally obtain post-processing image through IDA. In
this way, the security of the entire processing process is
guaranteed. The workflow of processing is as shown in Fig.
Figure 3. The architecture of mobile agent security service. 4.

In Fig.3 every part is described as follows:

• UDDI (Universal Description Discovery and
Integration). It is a web-based and distributed
implementation standard for provision information
for WebService. UDDI provides a set of standards-
based specifications for service description and
discovery, but also provides a set of Internet-based
implementations.
• RA (Registration Agent). In Agent systems, in order
to quickly distribute real-time consultation between
Figure 4. The workflow of security service.
the Agents, the function and location information of
each agent need to be known. So every agent need to
register in RA as so to prompt positioning of internal C. Experiment and analysis
system resources, accelerating decision-making In this paper, we do some experiment of web processing
process. of remote sensing images. Fig. 5 is shown. Fig. 5(a) and Fig.
• UIA (User Interface Agent). UIA provides a good 5(b) are the post-encrypted images at client side by IEA,
view and easy-handled interfaces for the images which have different encryption strength degree. Of course,
security services. The client can select the image may be encrypted in way as file is encrypted. The
encryption/decryption parameters for the web post-encrypted image is transmitted to server side through
processing images, including encryption algorithms, network. And then image is decrypted at server side, as
image format type and other information. shown in Fig. 5(c). The server continues to process image
• GA (Gateway Agent). It aims to accept and process according to the client request, for example in Fig. 5(d)
the corresponding Web services client-initiated. doing edge extraction for image. At last, the server side
• IEA (Image Encryption Agent), IDA (Image encrypts the post-processing image with different degree
Decryption Agent), IOA (Image other Agent). IEA according to security level and transmit it the client side. The
and IDA are the core of image security service. IEA client side decrypts the image through the IDA and obtain
achieves the image encryption service in order that post-processing image.
the security of web processing image can be
guaranteed through network transmission. IDA is
similar to IEA. IOA represents other agent services.
IEA and IDA are realized through mobile agent.
• NMA (Network Management Agent). It is mainly, in
distributed environment, integration and
management of various network protocols to deal
with network information exchange of
heterogeneous network, eliminating communication
(a) (b)

285
290

Authorized licensed use limited to: JADAVPUR UNIVERSITY. Downloaded on August 14,2023 at 10:15:27 UTC from IEEE Xplore. Restrictions apply.
 briefly discussed. The agent system configuration and the
concrete realization of the flexibility need to for further
research.
ACKONWLEDGEMENT
The work was supported by the National High-tech
R&D key Program (863 Program) (Grant No.
2007AA12020204), the National Natural Science
Foundation of China (Grant No. 40871200), and the PhD
(c) (d)
thesis independent research Fund of Wuhan University.

REFERENCES
[1] Z F Shen, J C Luo, etc., “Distributed computing environment for
remotely sensed image processing,” Computer Engineering and
Design, vol. 26, no. 1, pp. 1-4, 2005.
[2] Lee M C, “Semantic model-driven approach of web service systems,”
2009 International Conference on New Trends in Information and
Service Science, pp. 427-432, 2009.
[3] Hwang SY; Lim E P, etc., “Dynamic Web service selection for
reliable Web service composition,” IEEE Transactions on Services
(e) (f) Computing, vol. 1, no. 2, 2008.
Figure 5. Experiment results. [4] Shiguo lian, zhongxuan liu, “Secure media content distribution based
on improved set-top box in iptv,” IEEE Transactions on consumer
electronics, vol. 54, no.2, pp. 560-565, 2008.
IV. CONCLUSION AND FUTURE WORK [5] Jin Liu, Zhe Wu, et al., “Network distribution of remote sensing
At present, WebService goodly figure out web images based on multi-rank security,” the Fourth International
Conference on Computer Sciences and Convergence Information
distributed processing of remote sensing images, but it does Technology, pp. 1101–1104, 2009.
not guarantee the security of the processing image in [6] Fabio Bellifemine, Agostino Poggi,Giovanni Rimassa, Developing
network transmission environment. To settle this problem, Multi-Agent Systems with JADE, Wiley, London, 2007.
this paper proposes a security service method based on [7] Liao yali, The Design and Implementation of Image-Retrieval
mobile agent and WebService, ensuring the web security Prototype System based on Mobile Agent, Wuhan: Wuhan
processing of remote sensing images. In this Paper, only University, 2004.
part of the agent, the construction and realization rules are

286
291

Authorized licensed use limited to: JADAVPUR UNIVERSITY. Downloaded on August 14,2023 at 10:15:27 UTC from IEEE Xplore. Restrictions apply.