A COLLABORATIVE SECURITY MECHANISM BASED ON

REPUTATION AND TRUST LEVEL IN PERVASIVE SYSTEMS

Mohammed Nadir Djedid

Department of Computing, University of Sciences and Technology of Oran, Oran, Algeria

Keywords: Pervasive Systems, Identity Protection, Identification/Authentication, Collaborative Security Mechanism.

Abstract: The emergence of network technologies and the appearance of new varied applications in terms of services

and resources, has created new security problems for which existing solutions and mechanisms are

inadequate, especially problems of identification and authentication. In a highly distributed and pervasive

system, a uniform and centralized security management is not an option. It then becomes necessary to give

more autonomy to security systems by providing them with mechanisms that allows a dynamic and flexible

cooperation and collaboration between the actors in the system.

1 INTRODUCTION

The rapid development of mobile computing has

given rise to ubiquitous information systems: the

user has at any time, access to the global network

regardless of location or time ("anywhere - anytime"

access).

The challenge of pervasive systems is in this

perspective, to provide methodological frameworks

and protocols to permit the reliable, relevant and

efficient use of these systems. (Girma, 2006).

But this new trend reveals new security problems

for which solutions and existing mechanisms are

inadequate, especially for the problems of

identification and authentication. In a highly

distributed and pervasive system, a centralized and

homogenous security management is not

conceivable. It then becomes necessary to give more

autonomy to security systems, providing them with

mechanisms allowing a dynamic and flexible

cooperation and collaboration between the actors in

the system.

This paper will be an overview of the main

existing security systems and compare their

effectiveness and their ability to meet the major

identified security constraints of pervasive systems

like: Decentralization, Interoperability and

Interaction, Autonomy, Transpareancy and

Proactivity, Trust management, Scaling, and Privacy

protection. Finally, a generic architecture of a

security mechanism based on reputation and trust

level will be proposed.

2 IDENTITY AND PRIVILEGES

MANAGEMENT SYSTEMS

2.1 Systems based on Identity

Management

2.1.1 Radius

In Radius (Rigney and Al., 1997) the user sends an

Access-Request containing his/her authentication

information, and sends it to the server. The server

processes the request locally if it recognizes the user,

otherwise, it acts as a RADIUS Proxy "or

intermediate" by transmitting it to another server.

2.1.2 LemonLDAP

Two connection modes exist in LemonLDAP

approach (Wiki.LemonLDAP, 2007). In the pull

mode, when a user wants to access a protected

application, the system asks the user’s name and

password. Thus, after a successful authentication,

the user is redirected to the resource that he/she

seeks. In agent mode, the authenticated user accesses

a menu containing all the applications on which he

has access permissions.

258

Djedid M..

A COLLABORATIVE SECURITY MECHANISM BASED ON REPUTATION AND TRUST LEVEL IN PERVASIVE SYSTEMS.

DOI: 10.5220/0003837902580261

In Proceedings of the 2nd International Conference on Pervasive Embedded Computing and Communication Systems (PECCS-2012), pages 258-261

ISBN: 978-989-8565-00-6

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

2.1.3 OpenID

Open ID (Recordon and Reed, 2006) permits to

federate unique authentications and share attributes.

It provides the ability to authenticate to multiple

sites using a unique identifier OpenID.

2.1.4 Liberty Alliance

Liberty Alliance (Alsaleh and Adams, 2006)

proposes to combine the requirements of strong

authentication (authentication of multiple attributes)

by respecting the user’s privacy.

Just like OpenID, Liberty Alliance allows the

users with a single account to access multiple

services from different providers, but under the

condition that they must belong to the same "circle

of trust".

2.1.5 Shibboleth

Shibboleth (Shibboleth Development Team, 2009) is

an authentication mechanism, which permits to

federate the identification and supply, as the

mechanisms presented above, two possible

applications: authentication delegation and sharing

attributes.

2.1.6 Ws-security

WS-Security (OASIS, 2004) is a security protocol

called "point to point", which is dedicated to the

message exchange of information between web

services. Based on a mechanism of security tokens,

it is associated with digital signatures to authenticate

messages. Security tokens provide the identity of the

message sender, which is proved by an

authentication mechanism.

2.2 Systems based on Privilege

Management

2.2.1 Akenti

Akenti (Thompson and Al., 1999) is an architecture

designed to provide security services in a completely

distributed environnement.

The strength of Akenti is the autonomy offered

to the user who has the right to negotiates access to a

resource, by using authorization certificates.

2.2.2 Permis

Permis (Chadwick and Otenko, 2003) includes a

mechanism of static authority delegation. Thus, each

actor defines trust authorities having the right to

assign roles. In addition, a new version of Permis

(2006), can delegate authority dynamically, by

creating a chain of delegation.

The spread of trust by the chain of delegation is

considered as a breakthrough in the Permis project.

It allows the extension of security policies, but

obliges the authorities delegated to describe

manually the trusted entities who can take advantage

of privileges by the delegation.

2.2.3 Cas

CAS (Pearlman and Al., 2002) is a protocol

dedicated to control management in virtual

organizations (VO) like grid computing. CAS

assumes the role of supreme authority of a virtual

organization and allows to manage resources and

users between organizations working together in a

common project.

2.2.4 Voms

Virtual Organization Management Service (Alferi

and Al. 2004) closely resembles the CAS. The major

difference lies in the authorization mechanism.

Indeed, like if in CAS, the attributes concerning the

list of roles and groups members of the VO are

stored in the voms server, the authorization rules are

presented in the resource, which obtains the power

to decide the user’s right.

2.2.5 O2O

O2O (Cuppens and Nora cuppens-Boulahia, 2006) is

a security system for building a VO from several

VPOs (Virtual Private Organizations). Like a VPN

(Virtual Private Network), an VPO creates a bridge

between two organizations.

The policy of access control uses the same

federation mechanism as Liberty Aliance, so that a

unique profile can be attributed to each member an

organization and can thereby take advantages of the

privileges with the organizations linked to this VPO

gateway.

2.2.6 Sygn

In sygn (Seitz, Pierson and Brunie, 2005)

permissions are defined in the form of certificates

stored at the owner. For the creation of such

permissions, no interaction with a centralized system

is necessary, which makes it one of the Sygn’s

strengths. Sygn also offers the possibility to define a

permission on a set of resources.

A COLLABORATIVE SECURITY MECHANISM BASED ON REPUTATION AND TRUST LEVEL IN PERVASIVE

SYSTEMS

259

2.2.7 Gaia Os

Gaia OS Security (Roman and Al., 2002)

authenticates the user through different devices and

protocols. A number between 0-1 is assigned to each

device after authentication, which represents a

measure of trust in the device or protocol.

The advantage of GAIA OS, and in contrast to

the mechanisms seen previously, is that the entity is

measured digitally and not binary.

2.2.8 Tacp

TACP (Giang and Al., 2007) (Trust-Based Access

Control Policy) uses the concept of reputation

evoked earlier. The proposed approach begins by

estimating the value of trust that can be given to the

request sent by the user. In practice, each user is

assigned a confidence value between 0-1. Similarly,

each resource is a confidence level also included in

the interval [0,1]. Thus, if a user has a confidence

level higher than the confidence level of the resource

requested, he/she will be allowed, otherwise the

application will be rejected.

3 SYNTHESIS

This section will be dedicated to the comparison

between the different approaches presented above, in

relation to the major security needs of pervasive

systems. The result of the comparison is summarized

in two tables. The following two mentions are used:

 Y: Yes the need is supported by the solution.

 N: No the solution is not adapted for this need.

Table 1: Comparison of systems based on identity

management.

Constraints

Systems based on identity management.

RADIU

Lemon.

OPENI

Liber. Shib. WS

Decentralization

Y N Y Y Y Y

Interoperability

Y N Y Y Y Y

Trust spread

Traceability

Y Y Y Y Y Y

Autonomy

Y N Y Y Y Y

Transparency and

proactivity

Y Y Y Y Y Y

Flexibility

N N N N N Y

Privacy Protection

N N Y Y Y N

Scaling

Y N Y N N Y

Table 2: Comparison of systems based on privilege

management.

Constraints

Systems based on privilege management.

ent

Per

mis

Syg

IA.

Decentral.

Y N Y N

Y Y Y Y

Interoperab.

Y Y Y Y

Y Y N N

Trust spread

N Y Y N

N Y Y Y

Traceability

N Y N N

N Y Y N

Autonomy

Y Y Y Y

Y Y N Y

Transparen./

proactivity

N N N N

N N Y N

Flexibility

N N N N

N N Y N

Privacy

Protection

N N N N

Scaling

N Y Y Y

Y Y N Y

4 A COLLABORATIVE

SECURITY MECHANISM

BASED ON REPUTATION AND

TRUST LEVEL

In order to fully exploit the concept of spreading the

trust to interconnect security systems of various

domains, we propose a generic architecture in which

different modules are shown in Figure 1:

Figure 1: A collaborative security mechanism based on

reputation and trust level.

PEP: Is the external interface of the architecture

through which pass all the information in the form of

certificates, it has a particular module called "Trust

indicator" and reflects the reliability level of the site,

the site's reputation, the number of links of trust with

other sites etc. All these information are used to

assign a trust level to the site. The interface can also

verify the veracity of the certificates exchanged

between the system and the outside thanks to the

controller of certificates, and provide certificates

commanded by the PDP, to be sent to entities (user,

device).

PDP: Allows filtering access to the system via

the trust controller, and deciding to establish or

PECCS 2012 - International Conference on Pervasive and Embedded Computing and Communication Systems

260

revoke the trust with the sites. It also allows

applying the security policy defined in the PAP

module.

PAP: Is the module where the access control

policy is defined.

PIP: Allows the capture of the user’s context

(device used, connection type etc.). It also maintains

a table of trusted sites updated by the trust controller

module of PDP.

5 CONCLUSIONS

Through this study we came to the conclusions that

the concept of the propagation of trust in a dynamic

way is not fully exploited to interconnect the

security policies in various fields. Reply to this lack

could bring us closer to our goal of protecting the

identity of the users, and this in "globalizing" the

SSO system across domains: a single sign-on (SSO)

would not only provide access to several domain

resources belonging to the user, but also the

resources of the areas of trust where the user goes,

without being forced to decline again its identity.

This would avoid to re-circulate the information of

identification / authentication at the risk that it

would be intercepted by a third party.

Therefore, we proposed a generic architecture,

setting up a collaborative security mechanism based

on reputation and trust level accumulated by each

domain towards its peers. This work is a first step on

designing our architecture, and the future works will

be focused on calculating the value of the trust level

by providing a function that calculates this value.

REFERENCES

Girma, B., 2006. Thesis: Accès et adaptation de contenus

multimédia pour les systèmes pervasifs. INSA Lyon.

Rigney, C., Rubens, A., Simpson, W., Willens, S., 1997.

RFC 2138: Remote Authentication Dial In User

Service (RADIUS).

Wiki.LemonLDAP,2007

http://wiki.lemonldap.objectweb.org.

Recordon, D., Reed D., 2006. OpenID 2.0: a platform for

user-centric identity management. In The 2

ACM

workshop on digital identity management, pp.11-16,

ACM Press, Virginia, USA.

Alsaleh, M., Adams, C., 2006. Enhancing Consumer

Privacy in the Liberty Alliance identity Federation and

Web Services Frameworks Workshop on Privacy

Enhancing Technologies, pp. 59-77, Cambridge, UK.

Shibboleth Development Team, 2009. Shibboleth Project,

http://shibboleth.internet2.edu/.

OASIS, 2004. Web Services Security Specification.

http://www.oasis-open.org/.

Thompson, M., Johnston, W., Mudumbai, S., Hoo, G.,

Jackson, K., Essiari, A., 1999. Certificate-based access

control for widely distributed resources. In SSYM 08’:

Conference on USENIX Security Symposium, p.17,

USENIX Association Berkeley, USA.

Chadwick, D., Otenko, A., 2003. The PERMIS X.509 role

based privilege management infrastructure. Future

Generation Computer Systems Journal, Vol 19, No 2,

pp.277-289.

Pearlman, L., welch, V., Foster, I., Kasselman, C., Tuecke,

S., 2002. A Community Authorization Service for

Group Collaboration. In POLICY 02’: 3

International workshop on policies for Distributed

Sustems and Networks, pp.50-59, IEEE Computer

Society, Washington, USA.

Alferi, R., Cecchini, R., Ciaschini, V., Dell’Agnello, L.,

Frohner, A., Gianoli, A., Lörentey, K., Spataro., F.,

2004. VOMS, an Authorization Systemm fort Virtual

Organizations. In European Across Grids Conference,

pp.33-40, Verlag, Spain.

Cuppens, F., Nora cuppens-Boulahia C., 2006. O2O:

Managing Security policy Interoperability with Virtual

Private Organizations. In ICISS 06’. 2

International

Conference on Information Systems Security, pp. 101-

115, Kolkata, India.

Seitz, L., Pierson, J., Brunie, L., 2005. Sygn : A certificate

based access control in Grid environnements. Tech

Rep. INSA Lyon, France.

Roman, M., Hess, C., Cerqueira, R., Ranganathan, A.,

Campbel, R., Nahrstedt, K., 2002. Gaia: A

Middleware Infrastructure for Active Spaces. In IEEE

Pervasive Computing Vol.1, No 4, pp. 74-83.

Giang, P., Hung, L., Lee, S., Lee, Y., Lee, H., 2007. A

flexible Trust-based Access Control mechanism For

Security and Privacy Enhancement in Ubiquitous

Systems. In MUE 07’, International Conference on

Multimedia and Ubiquitous Engineering, pp.698-703,

IEEE Computer Society, Washington, USA.

A COLLABORATIVE SECURITY MECHANISM BASED ON REPUTATION AND TRUST LEVEL IN PERVASIVE

SYSTEMS

261