Scholarly article on topic 'Toward an Access Control Model for IOTCollab'

Toward an Access Control Model for IOTCollab Academic research paper on "Computer and information sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Procedia Computer Science
OECD Field of science
Keywords
{IOTCollab / "Internet of Things" / Security / "Access Control ;"}

Abstract of research paper on Computer and information sciences, author of scientific article — Mehdi Adda, Jabril Abdelaziz, Hamid Mcheick, Rabeb Saad

Abstract The increase in Internet-connected physical devices offers new possibilities and opportunities. This Internet of Things (IoT) fosters the development of new platforms, services and applications that connect the physical world (represented by physical objects) to the virtual world (represented by the Internet). The work presented here proposes a study of role and attribute-based access control models that tackle the security concerns of our already developed data sharing framework. The framework introduced a formal theoretical model, the IOTCollab domain specific language, and an integrated development environment that implements this model. We have extended this framework by completing the formal theoretical model with access control capabilities.

Academic research paper on topic "Toward an Access Control Model for IOTCollab"

(8)

CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Procedia Computer Science 52 (2015) 428 - 435

The 6th International Conference on Ambient Systems, Networks and Technologies

(ANT 2015)

Toward an Access Control Model for IOTCollab

Mehdi Addaa*, Jabril Abdelazizb, Hamid Mcheickb, Rabeb Saadb

aMathematics, Computer Science and Engineering Dep. University of Quebec at Rimouski, Rimouski (Qc), Canada Mathematics and Computer Science Dep. University of Quebec at Chicoutimi, Chicoutimi (Qc), Canada

Abstract

The increase in Internet-connected physical devices offers new possibilities and opportunities. This Internet of Things (IoT) fosters the development of new platforms, services and applications that connect the physical world (represented by physical objects) to the virtual world (represented by the Internet). The work presented here proposes a study of role and attribute-based access control models that tackle the security concerns of our already developed data sharing framework. The framework introduced a formal theoretical model, the IOTCollab domain specific language, and an integrated development environment that implements this model. We have extended this framework by completing the formal theoretical model with access control capabilities. © 2015 The Authors.PublishedbyElsevier B.V.This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the Conference Program Chairs Keywords: IOTCollab; Internet of Things; Security; Access Control;

1. Introduction

The Internet of Things is a revolution in the future of computing and the Internet, it is promoting the concept of anytime, anywhere connectivity for anything. IoT, even in its early stages, has changed the way consumers and organizations interact with each other and with the environment around them1. Thus, this new paradigm changes and we believe will change more and more business models, technology investments, consumer experiences, and even the day-to-day life tasks.

* Corresponding author. Tel.: +1-418-723-1986 #1850; fax: +1-418-724-1879. E-mail address: mehdi_adda@uqar.ca

1877-0509 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of the Conference Program Chairs

doi: 10.1016/j.procs.2015.05.009

As IoT research and technology is not yet mature enough2, there is no standard and unique definition. Given that, many definitions have been given to capture the different aspects and the meanings of the IoT concepts. Yet, it is largely agreed that those concepts are different from the conventional Internet.

IoT is already in our daily life-tasks. For instance, a mobile device may be used to monitor the security, light or heating systems of a house. However, IoT is facing many challenges such as scalability, availability, manageability, and security while promoting openness and collaboration.

The search for a compromise between the openness that is necessary for the collaboration, and the control and the restriction required to have a secure system is a multidimensional problem3. Indeed, a system or infrastructure intended to enable collaboration, among people and devices, must target primarily the easiness and the transparency, whereas the security aspect of that same system seeks the privacy, the integrity and an adequate authenticity. In this article, we deal with this dilemma. We tackle the access control problem and the theoretical means to integrate it to an already developed IoT collaborative framework.

The Role Based Access Control (a.k.a. RBAC) was formalized by Ferraiolo and Kuhn4 in 1992. Nowadays, the RBAC is still a predominant and a base model for advanced access control systems. First and foremost, this model was designed to overcome the burden of traditional ACLs by reducing the cost of access management. Attributes Based Access Control (a.k.a. ABAC)5 is another access control model. It uses direct proprieties associated with the subject, as well as with the resources and the environmental properties to grant rules to subjects. Another access control model to mention is the Task based Access Control (a.k.a. TBAC)678 where access control is modelled from the perspective of tasks. It is designed for the active security required by agent-based distributed computing and for workflow management. The authors assert that the permissions are managed with the purpose of being activated only in a just-in-time fashion, in addition to be synchronized with the processing of authorizations in progressing tasks.

In access control lists based mechanism, the service provider has to verify whether or not the subject is authorised to perform the requested operation on the requested objects. Another way to look at authorization assignment is via Capability Based Security. It is a security model in which the capability is a key-permissions relationship9. This relation could be seen as the relation between a car and its key. In other words, having the key means having the right and the permission to drive the car. From this perspective, the capability (the key in this example) is a sharable and unforgettable token of authority.

Many other access control models tackling the security issue in distributed computing and collaborative environments have been introduced. Either by extending or by redefining instance of existing models, the new access control models are focusing on specific issues or aimed at specific domains. Team-Based Access Control is one approach applying Role-Based Access Control in collaborative environments10. Context-Based Access Control 11 extends RBAC taking in account the notion of environment roles, and this, in order to support security in context-aware applications. Task-Role-Based Access Control is centred around both tasks and roles, and aims large commercial organizations and industrial companies12.

In light of this, this paper examines two of the main existing access control models applied to the Internet of Things in its second section. The third section, summarizes the features and characteristics of IoT, and the challenges access control models have to answer to in order to be suitable for IoT. The fourth and fifth section, motivated by previous sections, illustrate our vision of the Collaboration Role Based Access Control (CollRBAC) and the Collaboration Attribute Based Access Control (CollABAC) models for IoT. The sixth section of this paper is to be the integration of the resulting model to extend the formerly developed data sharing framework13. The seventh section, presents some concluding remarks and future work.

2. Access Control Model for IoT

The hype around IoT is receiving more attention recently. Although, it is not a new idea. At the end of the last century and the beginning of the current century, many attempts have been made to connect physical objects to the computer networks.

Later, different approaches have been proposed to integrate physical objects to the Internet. An information sharing architecture for IoT is presented in14. The authors suggested the concept of a user-centric architecture of the IoT that seamlessly integrates IoT objects, Web protocols, Web applications, and Social platforms, etc. In order to avoid connecting physical objects directly to the Internet, some approaches suggested abstracting those objects as services

by adopting the Service Oriented Paradigm15-17. For instance, the work presented by Guinard et al. in 15 describes the architecture of the Web of Things (WoT) based on the principles of the traditional Web such as scalability and modularity. They promote the reuse and the adaptation of existing Web technologies such as REST architectural style18 to interact with IoT objects.

Similar to conventional infrastructures, the main function of access controls is to guaranty right rights to the right subject on the right IoT object. To prevent unauthorized access to the Secure Discovery Service (SecDS) of their search engine, 19 proposed an extended attribute-based access control model to protect information belonging to different companies through different policies. In the same vein, Kerschbaum20 proposed an access control model for mobile physical objects based on the ABAC model. This later extends the attributes access control model to include the information about the trajectory of an object in supply chains. Thus, a trajectory-based policy has been integrated to provide a mutual access authorization and control.

Extending the role based access control model was claimed by 21 to enhance the security in service-based IoT infrastructure. The paper introduced the incorporation of contextual information in RBAC as a way to produce a better mechanism for access control in the Internet of Things. Following the same vision, and from a service-oriented perspective, 22 proposed a workflow-oriented and attribute-based access control model to treat access control issues within IoT. Attributes related to the subject, resources, the environment, and the task to have authorization for, all those parameters have been taken into consideration to obtain a fine-grained model.

Liu et al.23 proposed a feasible authentication and an access control model for the IoT. The adopted access policy inherits from the RBAC mechanism, while the Elliptic Curve Cryptography keys founded the authentication process.

Authors in 24 based their suggested access control model on devices capability and identity. The Identity Authentication and Capability based Access Control (IACAC) scheme creates the capability based on the identity to grant access on local network. This scheme still not fully suitable for small devices within the IoT. Following the same vein, 25,26 promoted the use of capability-based security approach to manage access control in the Internet of Things. Indeed, a capability defines the resources, the subject and the granted rights and authorisations. Key features supported by the Capability Based Access Control (CapBAC) include delegation and revocation of capability, as well as information granularity and standard capability representation through XML-based languages.

In 27, the authors proposed a model that combines location and time with security level to control access to the information within the IoT. The model is named Location-Temporal Access Control Model (LTAC). LTAC is meant to give access to requested operations on a defined node only if the requesting node is located in an appropriate location within the appropriate time interval regarding the object. In addition to the context of thing subject to the access demand, Oh and Kim28 included the identity and the internet address of the requester to the process of access control. Considering the web of things and REST-compliant resource-oriented web characteristics, they provided a decentralized access permission control structure

3. IoT, features and challenges of an access control model

In this section we present our principal considerations in designing the access control models presented in this paper. To ensure that the end result fits the requirements of the Internet of Things, we precede with the features and characteristics of this network of things, followed by the challenges an access control model should meet and have to deal with in such infrastructure.

3.1 IoT Characteristics

The Internet of Things, beyond the dilemma of its definitions, is a sub-layer or even another Internet. It represents a promise for the technology future and show some common features:

• The Scale: first thing to come to mind is the huge number of actors within IoT. Indeed billions of devices are already deployed. Thus, performing tasks over this set of objects makes the coordination process nearly impossible due to many constraints such as memory, energy, time and etc.

• Dynamic environment: IoT by nature is a dynamic network where actors are continually deployed; some new objects joining the network while others leaving.

• Massive amount of data: the management of the data sent and/or received by each object rises many issues including those related to Security and Privacy.

• High heterogeneity: mechanisms treating the interoperability dimension of the problems related to IoT's heterogeneous objects must be provided.

• Self-organized: the scale of IoT and its dynamic nature imply that, in this kind of network, the failure of an actor must have lowest or no regression effect.

• Limited Energy: a large share of IoT objects are tiny devices with limited resources, and they are designed to work with minimal energy consumption.

• Routing algorithm: the IoT communication is characterized by short interactions between devices. Those small communications aim to not produce any processing power overhead and to support the dynamic nature of IoT.

3.2 Access Control in IoT, the Challenges

As indicated in previous section, the Internet of Things is a demanding environment. Therefore, the access control model to be envisaged to collaboration in such environment must face and manage many challenges, such as:

• The huge number of connected systems and devices.

• The dynamism of IoT devices makes access control policies highly complex.

• The access control has to be suitable for groups and fine-grained access.

• The access control management mechanism has to be flexible.

• In addition, the mechanism has to support restrained-resources and simple devices.

• Finally, to provide a suitable easy to use interfaces for both consumers and devices needs.

4. CollRBAC: IoTCollab Role-Based Access Control Model

Fig. 1. Collaborative Role-Bases Access Control Model.

CollRBAC assigns role to users via the Role Assignment application Au, this operation may be seen as the first step in the authorization process. The second step in the other hand assigns a Permission to a given Role (Fig. 1).

Definition 1. (Permission) Given U0loT the universe of all IoT objects, Us the universe of all services, and U0ps the universe of all operations, a permission is a triplet < 0;, Sit 0Pt > such that:

• °t e UoIoT

• St £ Us

• 0Pt e u0ps

The universe of all permissions is denoted by UPerm.

Definition 2. (Role) Given UPerm the universe of all permissions, a role R is a finite set of permissions. Said differently R = [Permi\Permi £ UPerm}. The universe of all roles is represented by URole.

A role may be assigned to one or many users. The assignment relationship is defined bellow (See Definition 3).

Definition 3. (Role Assignment) Given URole the universe of all roles and Uy the universe of all users, the user-role assignment is a non-injective and non-surjective application Au from Ujj to URole.

Given the previous definitions, and for the sake of simplicity, here are some of the policies used to perform the basic functions of the CollRBAC (See Fig. 1) regarding access control features:

(1) addPerm associates a set of permissions [Perm} to the corresponding role R within the framework: addPerm{R,[Perm}y. V Permt £ [Perm}, Permt £ UPerm AR = R U [Perm} (1)

(2) rmPerm detaches a permission Perm; from a given role:

rmPerm(R,Permt). Permt £ UPerm AR = R — [Permt} (2)

(3) the Au from Uu to URole is the user-role application, it assigns a set of roles [R} to the appropriate user Ut: assigRole(R, Ut). UteUv ARE URole. Ut = Ut U [R} (3)

5. CollABAC: IoTCollab Attributed-Based Access Control Model

User Attributes Context Attributes Service Attributes

j Rsd. 5. atts x Rsd, WS{t\. atts x Ctx, atts \

Fig. 2. Collaborative Attribute-Bases Access Control Model.

Upon the reception of a service request Rsd (Fig.2), an access permission is demanded to allow or deny access to one or more IoT objects. When an access request is made, Attributes and Access Control Rules are evaluated by the Collaborative Attribute-Based Access Control mechanism to provide access control decision.

Definition 5. (User) Given Uythe universe of all users. A user U( is represented by a set of Attributes defining its identity and characteristics.

Definition 6. (Service) Given Us the universe of all IoT services, a service St is represented by a set of attributes such as data type, frequency.

Definition 7. (Context) Context is the set of attributes describing the state of the environment, the user and the service subject of the current demand. Contextual attributes include location and time.

Definition 8. (Access Control Rule) Given a service request Rsd = < S, IDs > from URsd, and the context of this request Ctx, the access control rule determines if the user who sent the request has the right to access the service S. This function, denoted by C(), returns a Boolean value that is equal to true when the access is granted, otherwise the value is equal to false. This function is formalized as follows:

€: Rsd. S. atts x Rsd. IDS[ 1]. atts x Ctx. atts ^ [true, false} (4)

6. Evaluation of the proposed models

6.1 Discussion: CollRBAC VS. CollABAC

In this section we compare both proposed models in order to validate their respective suitability for the collaboration and data sharing model IOTCollab presented in our previous paper 13. Table 1 illustrates this comparison against a set of criteria relevant to collaboration features and characteristics within IoT, see section 3.

Table 1. CollRBAC vs CollABAC

Criterion

CollRBAC

COllABAC

Least privilege principle Separation of duties Scalability

Scalable to a certain extent. With the growth of actors in the collaborative network, the huge number of objects and services may lead to an explosion of roles.

Providing subject with attributes may have an overload on the framework. Services and context attributes are basic building blocks. Thus, no specialized mechanism are to be deployed for this purpose.

The active nature of the ABAC makes it able to handle the dynamism of a collaborative system.

Dynamism support

In relation with the scalability criterion, the constant movement of actors in the network may lead to an overload on the access-roles management process.

Does not considerate contextual information in the decision making mechanism.

Contextual information

The Context attributes provide a fairly representation of the contextual information.

Granular

Low: lacks the ability to specify a fine- High through attributes representation. grained control on individual users in certain roles and on individual object instances.

Flexibility Low, regarding the responsiveness to High due to its high granularity.

the environment.

Active / Passive Passive Active

In order to validate the models, Table 1 evaluates the proposed access control models against the mentioned criteria. Both models are adapting features of the trusted and the largely-used RBAC and ABAC. Thus, they support the well-known principles of Least Privilege and Separation of Duties. From a collaborative perspective, the fact that access rules assignment is an application between groups of users on a set of objects is not fully sufficient. Often, a service in an instance of an actor might need specific permissions on an instance of an object at a particular time interval during the collaboration. The CollABAC provide a high level of fine-grained control over the Role representation of the CollRBAC model. In addition, one of the most important characteristics of any collaboration is the context. Contextual information are in the core of the IOTCollab model and fits with the attributes representation of the CollABAC.

6.2 The Collaboration Access Control Policy Process

As introduced in our previous work13, the formal data sharing model intended to ease the collaboration and data sharing in the IoT. For service discovery and delivery, the model relayed on a propagation query-response model and a straightforward whitelist/blacklist policy for access control. The model has been extended to embrace the new adapted access control model CollABAC.

The service pre-selection process in the data sharing model is founded over the satisfaction of the following conditions:

Given a service request s1 and service response s2.

• s1 is located near s2: the distance that is separating the two geographical points of those services is smaller or equal to a user fixed threshold,

• s1.data.d.t = s2.data.d.t,

• sLdata.d.u = s2.data.d.u,

• sl.data. frq.start > s2.data. f rq.start,

• sl.data.frq.end < s2.data.frq.end,

• s1.data. f rq.crn Q s2.data. f rq.crn (the frequency of s1 is covered by that of s2),

• Vop1 Es1.data.ops, 3op2 Es2.data.ops such that: op1.attribute = op2.attributeAopLvalue(Q V =)op2.value,

• VopL Es1.data.ctx.ops, 3op2 Es2.data.ctx.ops such that: opl.attribute = op2.attribute A opl.value (Q V =) op2.value.

The rule function C in the CollABAC access control model has been added to the set of conditions. Thus, a valid service offer is an offer that matches the service request and in which the associated service grants access to the demanded object.

C: s2. S. atts x si. 7DS[1]. atts x Ctx. atts ^ [true} (5)

Said differently, if the function in (5) returns true than grant subject si. 7D5[1] access to the service s2. S. Otherwise, if the function returns false, the opposite result will occur and the pre-selection of the service fails.

7. Concluding remarks and Future Work

In this article, we provided a brief but comprehensive-comparison study of authorization mechanisms for our previous formal data sharing model IOTCollab. We first presented access control requirements for IoT collaboration. Next, we proposed the CollRBAC and the CollABAC models to be evaluated in light of the IoT and IOTCollab requirements.

As IoT devices are generally resource-constrained, we plan to continue the real world experiments to test the effectiveness and performance of the proposed model; especially the potential network overload which may be induced by the propagation strategy and the access mechanism. To ease the integration of those concepts to a data sharing system, the dedicated domain specific language is intended to be extended along with the IDE to support it.

References

1. ITU. ITU Internet Reports, The Internet of Things. (2005). at <http://www.itu.int/wsis/tunis/newsroom/stats/The-Internet-of-Things-2005.pdf>

2. Perera, C., Zaslavsky, A., Christen, P. & Georgakopoulos, D. Context Aware Computing for The Internet of Things: A Survey. 1-41 (2013). at <http://arxiv.org/abs/1305.0982>

3. Tolone, W., Ahn, G.-J., Pai, T. & Hong, S.-P. Access control in collaborative systems. ACM Comput. Surv. 37, 29-41 (2005).

4. Ferraiolo, D. F. & Kuhn, D. R. Role-Based Access Controls. 554-563 (1992).

5. Hu, V. C. et al. Guide to Attribute Based Access Control (ABAC) Definition and Considerations. (2014). doi:10.6028/NIST.SP.800-162

6. Thomas, R. K. & Sandhu, R. S. Towards a task-based paradigm for flexible and adaptable access control in distributed applications. Proc. 19921993 Work. New Secur. Paradig. - NSPW '92-93 138-142 (1993). doi:10.1145/283751.283810

7. Thomas, R. K. & Sandhu, R. S. Conceptual foundations for a model of task-based authorizations. in Proc. Comput. Secur. Found. Work. VII 66-79 (IEEE Comput. Soc. Press, 1994). doi:10.1109/CSFW.1994.315946

8. Thomas, R. K. & Sandhu, R. S. Task-based Authorization Controls ( TBAC ): A Family of Models for Active and Enterprise-oriented Authorization Management. in Proc. IFIP TC11 WG11.3 Elev. Int. Conf. Database Secur. XI Status Prospect. 166-181 (Chapman \& Hall, Ltd., 1998). at <http://dl.acm.org/citation.cfm?id=646115.679940>

9. Levy, H. M. Capability-Based Computer Systems. (Digital Press, 1984).

10. Thomas, R. K. Team-based access control (TMAC): a primitive for applying role-based access controls in collaborative environments. in Proc. Second ACM Work. Role-based access Control - RBAC '97 13-19 (ACM Press, 1997). doi:10.1145/266741.266748

11. Covington, M. J. et al. Securing Context-Aware Applications Using Environment Roles. 10-20 (2001).

12. Oh, S. & Park, S. Task-role-based access control model. Inf. Syst. 28, 533-562 (2003).

13. Adda, M. & Saad, R. A Data Sharing Strategy and a DSL for Service Discovery, Selection and Consumption for the IoT. Procedia Comput. Sci. 37, 92-100 (2014).

14. Uckelmann, D., Harrison, M. & Michahelles, F. in Archit. Internet Things (Uckelmann, D., Harrison, M. & Michahelles, F.) (Springer Berlin Heidelberg, 2011). doi:10.1007/978-3-642-19157-2

15. Guinard, D., Trifa, V. & Wilde, E. A resource oriented architecture for the Web of Things. in 2010 Internet Things 1-8 (IEEE, 2010). doi:10.1109/I0T.2010.5678452

16. Guinard, D., Trifa, V., Mattern, F. & Wilde, E. in Archit. Internet Things (Uckelmann, D., Harrison, M. & Michahelles, F.) (Springer Berlin Heidelberg, 2011). doi:10.1007/978-3-642-19157-2

17. Pascual-Espada, J. Service Orchestration on the Internet of Things. Int. J. Interact. Multimed. Artif. Intell. 1, 76 (2012).

18. Fielding, R. T. & Taylor, R. N. Principled design of the modern Web architecture. ACM Trans. Internet Technol. 2, 115-150 (2002).

19. Shi, J. et al. SecDS : A Secure EPC Discovery Service System in EPCglobal Network. 267-274 (2009).

20. Kerschbaum, F. An access control model for mobile physical objects. Proceeding 15th ACM Symp. Access Control Model. Technol. -SACMAT '10 193 (2010). doi:10.1145/1809842.1809873

21. Zhang, G. & Tian, J. An extended role based access control model for the Internet of Things. in 2010 Int. Conf. Information, Netw. Autom. V1-319-V1-323 (IEEE, 2010). doi:10.1109/ICINA.2010.5636381

22. Zhang;, G. & Liu, J. A Model of Workflow-oriented Attributed Based Access Control. Int. J. Comput. Netw. Inf. Secur. 3, 47-53 (2011).

23. Liu, Y., Xu, K. & Song, J. A Task-Attribute-Based Workflow Access Control Model. in 2013 IEEE Int. Conf. Green Comput. Commun. IEEE Internet Things IEEE Cyber, Phys. Soc. Comput. 1330-1334 (IEEE, 2013). doi:10.1109/GreenCom-iThings-CPSCom.2013.231

24. Mahalle, P. N., Anggorojati, B., Prasad, N. R. & Prasad, R. Identity Authentication and Capability Based Access Control ( IACAC ) for the Internet of Things. J. Cyber Secur. Mobil. 1, 309-348 (2013).

25. Gusmeroli, S., Piccione, S. & Rotondi, D. IoT Access Control Issues: A Capability Based Approach. in 2012 Sixth Int. Conf. Innov. Mob. Internet Serv. Ubiquitous Comput. 787-792 (IEEE, 2012). doi:10.1109/IMIS.2012.38

26. Gusmeroli, S., Piccione, S. & Rotondi, D. A capability-based security approach to manage access control in the Internet of Things. Math. Comput. Model. 58, 1189-1205 (2013).

27. Lee, C., Guo, Y. & Yin, L. A Location Temporal based Access Control Model for IoTs. in AASRI Procedia 5, 15-20 (Elsevier B.V., 2013).

28. Oh, S. W. & Kim, H. S. Decentralized access permission control using resource-oriented architecture for the Web of Things. in 16th Int. Conf. Adv. Commun. Technol. 749-753 (Global IT Research Institute (GIRI), 2014). doi:10.1109/ICACT.2014.6779062