Trusted Platforms to secure Mobile Cloud Computing - Semantic Scholar

Trusted Platforms to secure Mobile Cloud Computing Thinh LE VINH

Samia BOUZEFRANE

CEDRIC Lab Conservatoire National des Arts et Métiers - CNAM Paris, France [email protected]

CEDRIC Lab Conservatoire National des Arts et Métiers - CNAM Paris, France [email protected]

Abstract— Taking full advantage of the availability of Cloud computing facilities, Mobile Cloud Computing (MCC) is becoming more and more popular as an element of mobile technology. This availability enables mobile users to use the cloud infrastructure to overcome the limitations of mobile technology; namely limited data storage, processing power and battery life. With the increasing use of mobile cloud computing, security issues remain a challenge. This paper investigates some available trusted platforms in order to pinpoint the question on how to secure the MCC under trusted platforms. This investigation comprises an overview of the security aspects of Trusted Platforms including Secure Elements (SE), Host Card Emulation (HCE), Trusted Execution Environment (TEE), and Trusted Platform Module (TPM). In addition, it shows how these platforms can be suitable for securing mobile Cloud computing. Index Terms—mobile computing, trusted platforms, cloud computing, security.

I. INTRODUCTION According to Cisco [1], “about 36 percent of worldwide mobile service revenues come from data; by 2016, it will increase to 46 percent. Similarly, the 1.3 exabytes per month of mobile data traffic generated today is expected to increase eight-fold by 2016”. Mobile devices will more and more deal with heavy computation tasks because principally of the huge size of data that will be more and more manipulated. Besides, secure elements growth will reach 64% at the end of 2014, according to EuroSmart1, with 435 millions of units, while smart secure contactless devices will exceed one billion of units. With the emergence of new information and communication technologies, smart devices combined with communication infrastructures such as Internet and mobile networks need to provide reliable services. For this purpose, many application domains with high social and business impact such as personal healthcare, home automation, mobile payment, may use mobile devices that rely on trusted environment. This trusted environment may be provided or based on many hardware or software platforms such as SIM cards, secure elements for Near Field Communication (NFC), Host Card Emulation (HCE), TEE (Trusted Execution Environment), or TPM (Trusted Platform Module). Even if these trusted platforms are extremely constrained in terms of energy, computational power and memory, they are designed to solve major constraints of 1

http://www.eurosmart.com/publications/market-overview

security. In this article, we will investigate the features of these trusted platforms and demonstrate that they are suitable to Mobile Cloud Computing (MCC) context. In fact, in MCC, the trend nowadays is to outsource data and computation of mobile devices to the Cloud in order to extend the limited-resource mobile devices and to benefit from unlimited Cloud capacities such as storage. However, MCC must offer mechanisms to guarantee data protection and security in untrusted environment. The contribution of this paper is to look into the Trusted Platforms features. Section 2 provides a summary in short of Trusted Platforms. Section 3 deliveries the features of trusted platforms in comparison. Section 4 discusses the different architectures that may implement a MCC model. Section 5 discusses how MCC security issues can require the use of trusted platforms. Section 6 concludes the paper with some perspectives. II. TRUSTED PLATFORMS The emergence of a new generation of mobile devices leads to the growing number of security concerns demand attention. Mobile users use their smart-devices for a variety of application in an open environment. They are provided the reliable services to reduce the risks associated with the security threats. As a result, mobile devices need Trusted Platform to meet a proper implemented security. This Trusted Platform can be a Secure Element, a Trusted Execution Environment, a Host Card Emulation, or a Trusted Platform Module. A. Secure Element A smartcard is a tamper resistant micro-controller [2] whose security is enforced by multiple software and hardware countermeasures. A secure element (SE) is a smart card embedded in an NFC device such as NFC-enabled smart phone, able to store sensitive data such as PIN code and security keys and perform internal computation such as cryptographic processing when needed by the terminal (phone). A SE has the following physical features: 8, 16 to 32 bits of CPU, 32 to 64 KB of PROM that hosts the operating system, and 1 to 4 KB of RAM to execute programs, 1KB to 128 KB of EEPROM/Flash to store data and applications, while a dedicated crypto-processor is used for cryptographic computation. To interact with a SE, the terminal sends APDU commands and the SE responds using APDU responses. The SE is generally a Java Card platform (see Fig. 1) composed of a Java Card Virtual Machine (JCVM), a Java Card API and a Java Card Runtime Environment (JCRE).

JCRE implements the Java Card mechanisms that are intrinsic to the smart cards such as transaction management. When the SE is a SIM card, it is enriched with Java Card packages that allow interaction with the Mobile Network Operator (MNO) such as sending SMS, generating a phone call, etc.

NFC controller to perform the communication with the contactless smartcard reader infrastructure.

Fig. 3. NFC mobile device

Fig. 1. Java Card platform

In addition, Java Card platforms are designed in respect to Global Platform specifications [3]. In other terms, each Java Card platform is composed of isolated E²PROM areas called security domains (SD) as in Fig. 2. A SD is assigned to one service provider (SP) and may contain one or several applications belonging to the same SP. Each Java Card application is composed of one or several Java Card applets, generally called Cardlets. Controlling Authorities' Security Domain(s)

Global Services Application(s)

Application Providers' Security Domain(s)

Application Provider Application(s)

Issuer's Security Domain

Card Issuer Application(s)

GP API

B. HCE “Card Emulation” has been initially defined by the NFC Forum [4] and integrated as a part of the core set of NFC specifications. HCE (Host Card Emulation) allows the software emulation of a smart card-based application. Prior to December 2013, HCE was deployed firstly in Blackberry OS. Since then, the feature has also been supported by Android OS, following the launch of version 4.4, codenamed “KitKat”. It provides an API set that helps developers to control the NFC interface and send commands to NFC-enabled devices. In this case, the data is routed to the host CPU on which Android applications are running directly, instead of routing the NFC protocol frames to a secure element. HCE operation and protocols are shown in figure 4 as in [5].

RTE API

OPEN and GP Trusted Framework Runtime Environment

Fig. 2. Security Domains as defined in GlobalPlatform

As in Fig.3, each NFC device has a NFC controller that allows a contactless communication with respect to three modes: Read/write mode: the application running on the mobile device is able to read and write data on RFID tags according to an NFC Data Exchange Format defined by NFC forum [4]. Since 2010, Google proposed an API to implement this mode. Peer to peer mode: called Android Beam in Android platforms, this mode allows a tap of NFC devices together to exchange data between them. Emulation card: unlike the other modes, the emulation card mode involves the use of a SE and allows communication between the mobile and a contactless reader. In case SE is a SIM card (see Fig.3), a Single Wire Protocol (SWP) connects the SIM card with the

Fig. 4. NFC card emulation without a secure element HCE supports APDU stack protocol

Applications could therefore be developed to emulate any classic contactless smartcard application using the ISO 144434 standard, such as loyalty, access control, ticketing or payment applications. In addition, HCE is simply an ordinary Android application without any specific hardware or software-based security services. Security concerns relating to HCE are further compounded by industry reports indicating that Android is, by some distance, the most attacked of all mobile environments. According to [6], HCE is good for the NFC ecosystem as a whole; it will make NFC more accessible and versatile to developers, and more familiar to end-users, increasing mass

market adoption as a result. It is also best suited to use cases where the user's stored credentials are of low value and where stringent security requirements are not mandatory and the emulated NFC application is not based on direct implementation of a current, pre-existing card application. However, service providers evaluating HCE for payment and other high-value NFC services should proceed with caution. HCE presents a new raft of challenges and has the potential to diminish both the transaction security and the end user’s NFC service experience. HCE remains in its infancy. Until more OS providers support the HCE model, deployments will remain vulnerable to the prevailing challenges associated with global Android OS utilization. As a software solution, HCE is considered as less secure than a physical SE. For example, the HCE cannot handle counter-measures for physical attacks. Unlike secure devices, HCE concept does not target applications with high-level security, and relies on larger resources (memory & processor) because it uses the host OS. In addition, it is independent from SE providers. C. TEE The GlobalPlatform Trusted Execution Environment (TEE) [3] defines a standardized isolation environment for mobile devices in which sensitive code, data and resources are processed away from the main operation environment, software and memory on the device. This isolation is enforced by hardware architecture and the boot sequence uses a hardware root of trust in the system on chip package making it highly robust against software and probing attacks. In addition, code running in the TEE and using protected resources (known as “Trusted Application”) is cryptographically verified prior to execution, leading to high integrity assurance. Because it provides an isolated runtime environment entirely inside the mobile device, the TEE enables advanced device or peripheral security use cases such as securing the user interface, or controlling access to an NFC chip. As such, the TEE can be used as a distinct security coprocessor or provide a trusted “bridge” between the user and other security technologies such as secured user interface or OS user permissions on one side, and Secure Element access control on the other. The main operating system and rich applications then run as normal on the device, accessing the functionality of the Trusted Applications via a standardized ‘Client API’. Trusted Applications are written to an ‘Internal API’ which ensures portable trustworthy access to secure resources, cryptographic operations and secure storage regardless of the underlying SoC hardware. Fig. 4 shows the architecture of the TEE [7]. Furthermore, the TEE manages and executes trusted applications built in by device makers as well as trusted applications installed as people demand them. Trusted applications running in a TEE have access to the full power of a device's main processor and memory, while hardware isolation protects these from user installed apps running in a

main operating system. Software and cryptographic isolation inside the TEE protect the trusted applications contained within from each other like with ARM Trustzone platform [8].

Fig. 5. Architecture of GP TEE

Device and chip makers (such as ARM) use TEEs to build platforms that have trust built in from the start, while service and content providers rely on integral trust to start launching innovative services and new business opportunities. Hence, the Trusted Execution Environment (TEE) [8] is a secure area that resides in the application processor of an electronic device. Devices with a TEE will provide consumers with more secure, user-friendly experiences that simplify and speed up how they interact with their digital world. This will enable them to use their smart, connected devices more frequently to access an increasing range of applications and services in a secure way. This includes mobile payment, enterprise productivity and mobile banking applications, as well as online commerce and premium content services. SierraVisor and SierraTEE (ARM), ARM TrustZone, and Trustonic TEE are some examples of platforms implementing the TEE concept. The motivation of TEE is to serve as a combination between mobile-device OS (called rich OS) performance and SE security; because SE is resilient to attacks but limited by its resources and the rich OS is vulnerable to attacks. Fig 6 shows the shape of arrows including width and height to represent security and usability characteristics in Rich OS, TEE and SE. Taking “Attack Resistant” as a specific example, arrow’s height indicates that the Attack Resistant of the three is acceptable; the width shows that SE is the best and TEE, Rich OS are fine and poor respectively. With larger resources than a SE, the RAM size of a TEE may reach 1MB with a Flash memory shared with a Rich OS. TEE may cooperate with SE to guarantee a high level of security. D. TPM Unlike the platforms described below that are dedicated to mobile devices, the Trusted Platform Module (TPM) is a hardware platform dedicated to PCs, servers, personal digital assistant (PDA), printer, or mobile phone to enhance security

in an ordinary, non-secure computing platform and convert them into trust environments. TPM implements mechanisms and protocols to ensure that a platform has loaded its software properly. Independent from the host OS, it stores secret keys to encrypt data files/messages, to sign data, etc.

B. Hardware SE, TPM and some implementations of TEE have their own limited hardware resource. In contrast, HCE and TEE have lager resources because they have access to the hot resources. Thus, the TEE provides a more powerful processing speed capability and greater accessible memory space than SE. Unlike the other platforms, HCE has at its disposal all the memory resources of the Android platform, but HCE is not a piece of hardware. C. OS and API support At the moment, HCE is not isolated from the device's operation system and supported by Android OS in version 4.4. However, in the software form, HCE is considered less secured than others.

Fig. 6. Rich OS, TEE and SE positioning

Called Trusted Computing Platform Alliance until 2003, Trusted Computing Group (TCG) is a consortium of companies: Compaq, HP, IBM, Intel, Microsoft, AMD, etc. that design TPM specifications. The TCG requires three Roots of Trust in a trusted platform: Root of Trust for Measurement (RTM), Root of Trust for Storage (RTS), and Root of Trust for Reporting (RTR). It has recently released the TPM 2.0 library specification that provides a support for additional cryptographic algorithms, enhancements to the availability of the TPM to applications, enhanced authorization and management mechanisms. According to TCG, “specifications will detail how the TPM can be implemented in various platforms through TCG platform specific specifications. These future specifications include the TPM Software Stack specification (TSS) and separate specifications for PCs, mobile, embedded and virtualized platforms”. III. COMPARISON OF TRUSTED PLATFORMS FEATURES In previous sections, we presented the overview of available Trusted Platforms. In this section, we go to the comparison of the different features of Trusted Platforms. Table 1 summarizes our study of the Trusted Platforms. In general, almost security solutions for mobile computing rely primarily on SE. Nevertheless, in some cases, the solutions do not require very strict security requirements and the level of security required should be balanced to the importance of the context. In this paper, we discuss the following criteria. A. Form and Dependency Different from SE and TEE, HCE and TPM present only in the software and hardware form prospectively. The former is simply the android program where the user's stored credentials are of low value. The latter is a computer chip and considered as the cheapest way to enhance security in an ordinary, nonsecure computing platform. In term of Dependency, TEE works as the framework for mobile-device security. It is also known as a layer between mobile OS and typical SE.

D. Security features In this aspect, the SE is the highest security level for the mobile computing. However, it also required extra cost for the implementation. In some case, the TEE is sufficient for most application. Moreover, the TEE can work with another trusted platform to enrich the security level. For the moment, TEE standardization is not yet finalized. By contrast, HCE concept does not target application with high-level security. Security concerns relating to HCE are further compounded by industry report indicating that Android is, by some distance, the most attackable of all mobile environment [6]. E. Role The role of HCE is a blur in comparison with other trusted platforms. Some drawbacks deal with HCE such as low power mode, roaming and no data connectivity scenarios and transaction speed. As stated in [6], HCE still remains in the beginning stage and needs a long way to go before it can establish a certification schema that is comprehensive and robust enough to gain the confidence of service providers and end-users. F. Physical / Software attack The SE’s physical protection depends on the type of chips. Most of them have strong protection. TEE, SE and TPM are designed to prevent from unauthorized software attack, work well in many use cases including e-commerce, citizen to government applications, online banking, and other fields where there is high security level required. G. Privacy In order to be against the increasing of the security threats, the TPM supports covering privacy protection and interoperability on multi-platforms including PCs and Mobile devices. Hardware manufacturers can flexibly implement the architecture of TPM chips. The owner can also flexibly deploy this architecture depending on the security solutions. Regarding the software implementation of available trusted platforms, HCE, supporting strongly by Google, has gained some attention in the NFC technology. With SE-less, HCE faces with many security risks in comparison to SE and also requires

more attention from the developer side. However, it enriches some factors; namely development costs, time to market and the need to cooperate with other parties. In the next section, we investigate the different architectures of mobile cloud computing in order to show how to use the trusted platforms to secure them. IV. MOBILE CLOUD COMPUTING ARCHITECTURES To highlight the motivation for mobile cloud computing, different architectures in the literature [9, 10, 11, 12, 13] have been defined to cater to different use cases in mobile cloud computing. The major benefit of cloud computing for mobile devices is to enable running applications between resourceconstrained devices and Internet-based Clouds. Hence, resource-constrained devices can outsource computation/communication/resource intensive operations to the cloud. Mobile cloud applications based on mobile cloud collaboration may deploy their components into different places including local smart phone, virtual machines in cloud and cloudlets. The principal motivation of offloading is to achieve less execution time and less energy consumption within mobile devices. In the following, we describe the possible architectures for mobile cloud computing. A. Cloud computing with mobile devices Commonly, mobile cloud computing means to prolong the capabilities of storage/computation-limited devices, and to provide seamless access to data/application on a remote resource rich server from anywhere. The network connectivity from the device to the cloud server needs to be optimized to ensure the quality of service and seamless handover. B. Virtual cloud computing provider Another approach [14] is to build up a cloud with peer-topeer connected mobile devices for data storage and processing so that mobile devices are resource providers of a virtual cloud. Through Hyrax platform, the author in [15] demonstrates the possibility of building a cloud with mobile devices such as Android smartphones to provide basic functions and services. C. Cloudlets as Intermediate Offload Elements Similar to the concept of hotspots, the cloudlet concept proposed by Satyanarayanan [16] is another approach to mobile cloud computing. Cloudlet approach is an alternative solution to remote cloud server which would present latency and bandwidth issues. Indeed, a cloudlet is viewed as a multicore computer installed in the public infrastructure with connectivity to remote cloud servers. Hence, the cloudlet is used by the mobile device to offload its workload while ensuring low delay and high bandwidth. Simanta et al. in [17] proposed an example of implementation of the concept of cloudlets with the objective of enhancing processing and conserving battery power in the mobile devices notably in hostile environments where networks are unreliable. The key feature of the architecture is that offload elements are stateless. Communication between the cloudlet and the cloud core is done only during setup and provisioning. Once the cloudlet is provisioned, it works in a

disconnection mode with the core and in a connection mode with the mobile device. When a mobile device is connected to a cloudlet, an application overlay is offloaded from the mobile device to the cloudlet. An application overlay represents the difference between a base VM with only an operating system installed and the same VM with the application installed. The cloudlet may host multiple VMs. V. SECURITY ISSUES IN MOBILE CLOUD COMPUTING MMC not only takes advantage of the cloud computing facilities but also inherits the security threats of conventional cloud computing. Using the cloud for mobile devices raises a challenge of security and trust issues. The current challenges of MCC can be listed here such as malware, virus, privacy, financial fraud, content protection, enterprise data and secure space. Due to this matter, there is a rise of awareness and an acceleration of adoption of security solutions. In this section, to address the role of Trusted Platforms in MMC, we discuss two security concerns such as Security of the mobile devices and Security of the Cloud. A. Securing the mobile devices In terms of Security for mobile users, the available Trusted Platforms have played their security role well in mobile device environment. In addition, these trusted platforms can work together to meet the security requirement. For example, TEE can serve as a complementary security countermeasure and integrating nucleus for service that depends on partial identities distributed across SEs, such as Active Stickers, Secure Micro SD Cards, UICCs, and Embedded SEs. The TEE combines these into integrated solutions which assure seamless interaction and security of processes that are executed in the periphery of the respective [3]. The combination of TEE and TPM Mobile is another typical example. Because of the advantage in on-processor technologies such as the TEE combining with the flexibility of the TPM protocols, this means it is possible to implement the TMP as an integrated solution or in firmware. As a result, the TPM Mobile functionality is implemented as a Trusted Application in the TEE. Securing mobile devices means to secure mobile apps and data. Recent work has been devoted to design trust environment for mobile devices as in the following. - Security of mobile apps Since apps are executed in untrusted mobile devices, trusted platforms are essential to overcome security constraints. Depending on the security level that is required, apps may rely on physical or software trusted platforms. For example, Roland developed, in [18], a prototype that emulates secure elements within Android platforms. Such an open emulator is used for debugging and rapid prototyping of secure element applications. The emulator is a complete Java Card API implemented within Android. Comparable to HCE of Google, the emulated secure element allows security and trust, and is proposed to replace the physical SE only for long-term testing and for showcasing of applications.

authentication, confidentiality and integrity are provided to illustrate the TPM use. - Security of Data Confidentiality and integrity are required to secure data stored on the mobile device and to allow transfer through the Internet. Well-established protocols such as SSL/TLS should be used for data transfer while confidentiality and integrity may be computed by the trusted platforms of the mobile device since they integrate a crypto-processor with trust storage, and implement cryptographic algorithms. Depending on the data sensitivity, data may be encrypted or not and may be stored in a distinct storage memory. In fact, if the data is sensitive like a PIN code, a generated security key, or a password, it can be stored in the Flash memory or the E2PROM of the trusted platform (SE or TEE). If the data size exceeds the memory available in trusted platforms, data can be stored in the mobile device memory that is protected by the trusted platform like in TEE, or even outsourced if the Cloud provides a trusted environment. In case data is not sensitive, it can be stored in the untrusted mobile device or on the Cloud. B. Securing the Cloud - Security of the apps and VMs To secure apps, the authors, in [19], propose the concept of a Cloud of Secure Elements (CoSE) where the secure services are hosted by servers rather than by smartphone Secure Elements. They discuss the use of CoSE for mobile payments, and illustrate how an NFC smartphone may be efficiently used as a bridge between an NFC reader and an Internet server of secure microcontroller that hosts EMV applications. The advantage of this solution is to assign dynamically a physical secure element on the cloud for a sensitive application. In cloud computing, the service provider must achieve auditability, for attesting whether security setting of applications has been tampered or not. As stated in [20], auditability can be achieved using remote attestation techniques that require a trusted platform module (TPM). The TPM contains an endorsement private key that uniquely identifies the TPM. At boot time, the host computes its system state (boot sequence, namely the BIOS, the bootloader, and the software implementing the platform) as a sequence of hashes securely stored within the TPM. A remote party can use this sequence of hashes to authenticate the host and to trust the host platform. In traditional systems, the operating system is linked to the hardware machine regarding the attestation process. However, in a virtualized environment like the clouds, VMs can dynamically migrate from one location to another, making the remote attestation mechanism not sufficient. Santos et al. [21] propose to secure the hardware layer using hardware TPM, and secure VMs using a secure virtual machine monitors. VM migration is possible if both source and destination servers are trusted. In [21, 22], the authors designed efficient protocols for trust establishment and management. Patidar et al., in [23], propose a system in which cloud computing system is combined with trusted computing platform using a TPM. Security mechanisms like

- Security of Data Many authors propose security mechanisms for Cloud storage. For example, Hsueh et al. [24] proposed hybrid cryptography, and digital signatures to provide security requirements for data storage of mobile phones. Their solution is used to avoid malicious attackers from illegal access and to share desired information with targeted friends by distinct access rights. Unlike this kind of solution, the hardware-based encryption can take advantage of the TPM and does not require user intervention or impact system performance. For instance, [25] have published a Trusted Storage specification to provide a manageable, enterprise-wide means for implementing fulldisk encryption using TPM. This trusted storage notion may be extended to target the storage provided by the data centres of the Cloud. In terms of digital right management (DRM), Zou et al. [26] propose Phosphor, a cloud-based mobile DRM scheme with a subscriber identity module (SIM) card in mobile phone, to avoid illegal distribution and piracy of digital contents (like video, audio, e-book, etc.). The SIM card hosts a Licence State Word (LSW) protocol that decrypts each digital content received from the Cloud using the decryption key stored on the SIM card. The drawback of this solution is the sharing of a secret key between the Cloud and the SIM card. In other words, the Cloud has to maintain a key for each user. C. Authentication In the cloud environment, authentication and access control are more crucial than ever since the cloud services and the cloud storage may be accessible to anyone over the Internet. The TPM can easily provide stronger authentication than username and password. In [27], the authors present a protocol to authenticate a nearby Cloudlet by a mobile device that handles NFC applications. For authentication, the mobile device uses a SE and involves the Mobile Network Operator and a trusted service management that belongs to a Cloud computing. D. Data privacy In addition to data protection as presented above, privacy is a means to allow the user having a complete control on his personal data hosted on the Cloud computing. In some MCC frameworks, like in MobiCloud, Huang et al. [28] introduce the notion of Extended Semi-Shadow Image (ESSI) that is not exactly the same as a virtual image because an ESSI could be an exact clone, a partial clone, or merely an image that has extended functions of the physical device. CloneCloud [29] is another MCC platform that uses offloading technique to perform computing on a resourceful server. For this purpose, the authors use device clones, and assume that the clone VM is by default a trusted one, and mention establishing trust as future work. The drawback of the MobiCloud and CloneCloud platforms that rely on device clones on the Cloud, is that no privacy means is provided to the user to let him choose the personal data to outsource or not on the device clone. If a TEE

based solution is adopted for example in this context, all the personal data that the user does not want to outsource can be stored in a secure manner by the TEE that has larger memory storage than a SE. E. Trusted platforms based MCC architectures In this paragraph, we investigate the use of trusted platforms regarding each MCC model. In the virtual cloud that is based on peer-to-peer mobile devices, the trusted environment can be built on the trusted platforms (like SE, TEE and HCE) of the mobile devices. Depending on the required security level, the trust can rely on SE, SE combined with TEE, or HCE. These trusted platforms can be used to guarantee the integrity and the confidentiality of the information exchanged between the devices and can be used by the mobile devices for authentication and trust. In the Cloud server model, the trust environment can be built using the trusted platforms (SE/TEE) of the mobile device combined with trusted platforms on the Cloud such as the Cloud of SE (CoSE) proposed by [19]. However, this kind of trusted environment may rely on a global security strategy so that the involved trusted platforms hosted on the mobile and the Cloud may interact properly to achieve the same goal. In the Cloudlet architecture, we need to design two trust environments: one trust environment based on the Cloudlet and the Cloud and a second environment based on the mobile device and the Cloudlet. Using a TPM on the Cloudlet and SE/TEE on the mobile device allows implementing different security mechanisms and attesting the authentication of the different involved platforms. The TPM of the Cloudlet can also contribute to define a trust environment with the Cloud. Regarding the service model of cloud computing, trusted platforms like TPMs can be used to ensure trusted VM launch in an untrusted IaaS environment like in [30]. VI. CONCLUSION There is considerable consensus that the trusted platforms are helpful in Cloud security. In this paper, we have presented several Trusted Platforms architectures and described how the MCC security gains the support from the Trusted Platforms. By introducing the overview architecture of several available Trusted Platforms and Mobile Cloud Computing and making the comparison of these Trusted Platforms, we have presented our approach which is to build security and trust in this paper. The former focuses on either the security of mobile devices or the security on cloud enabled by trusted platforms. The latter concentrates to distributed trust environment for mobile devices and Cloud side including Cloud server model and Cloudlet model. We plan to design a complete trusted environment based on trusted platforms for mobile cloud computing and implement a fully functional prototype based on our design and evaluate its performance in the near future. REFERENCES [1] Henky Agusleo & Neeraj Arora, The Road to ‘Cloud Nine’ How Service Providers Can Monetize Consumer Mobile Cloud, White paper, Cisco Internet Business Solutions Group (IBSG), Feb. 2013.

[2] Jurgensen, T.M. ET. al., Smart Cards: The Developer's Toolkit, Prentice Hall PTR, 2002,ISBN 0130937304 [3] GLOBAL PLATFORM specifications:: http://www.globalplatform.org/specifications.asp [4] NFC Forum : http://nfc-forum.org/ [5] http://developer.android.com/guide/topics/connectivity/nfc/hce.h tml#HceServices [6] http://www.simalliance.org/ [7] Globalplatform TEE White Paper http://www.globalplatform.org/TEEevent/globalPlatform_tee_w hitepapers.asp [8] Trustonic: www.trustonic.com [9] F., Niroshinie, S. W. Loke, & W. Rahayu, "Mobile cloud computing: A survey." Future Generation Computer Systems 29.1 (2013): 84-106. [10] Xiaopeng Fan, Jiannong Cao and Haixia Mao “A Survey of Mobile Cloud Computing”, ZTE Communications, March 2011, Vol.9 No.1, pp. 8-12, ISSN 1673-5188. [11] Sweta Patel : A Survey of Mobile Cloud Computing: Architecture, Existing Work and Challenges Computer Engineering Department R.K. Uuniversity , India Volume 3, Issue 6, June 2013 ISSN: 2277 128X. [12] H. XuhuiLi and Y. Zhang, “Deploying Mobile Computation in Cloud Service,” in Proceedings of the First International Conference for Cloud Computing (CloudCom), 2009, p. 301. [13] Muhammad Shiraz, Abdullah Gani, “Mobile Cloud Computing: Critical Analysis of Application Deployment in Virtual Machines”, 2012 International Conference on Information and Computer Networks (ICICN 2012) IPCSIT vol. 27, pp.11-16, IACSIT Press, Singapore. [14] G., Huerta-Canepa, & D., Lee, "A virtual cloud computing provider for mobile devices.", in: Proceedings of the 1st ACM Workshop on Mobile Cloud Computing & Services: Social Networks and Beyond, MCS’10, ACM, New York, NY, USA, 2010, pp. 6:1–6:5. [15] G., Huerta-Canepa, & D., Lee, "A virtual cloud computing provider for mobile devices.", in: Proceedings of the 1st ACM Workshop on Mobile Cloud Computing & Services: Social Networks and Beyond, MCS’10, ACM, New York, NY, USA, 2010, pp. 6:1–6:5. [16] E. Marinelli,“Hyrax: cloud computing on mobile devices using MapReduce,”Master thesis, Carnegie Mellon University, 2009. [17] M. Satyanarayanan, P. Bahl, R. Caceres, N. Davies, The case for VM-based cloudlets in mobile computing, IEEE Pervasive Computing 8 (2009) 14–23. [18] Michael Roland, « Debugging and Rapid Prototyping of NFC Secure Element Applications”, in Mobile Computing, Applications, and Services Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering Volume 130, 2014, pp 298-313. [19] Pascal Urien, Selwyn Piramuthu, « Securing NFC Mobile Services with Cloud of Secure Elements (CoSE)”, in Mobile Computing, Applications, and Services Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering Volume 130, 2014, pp 322331. [20] Qi Zhang • Lu Cheng • Raouf Boutaba, « Cloud computing: state-of-the-art and research challenges”, The Brazilian Computer Society 2010, J Internet Serv Appl (2010) 1: 7–18.

[21] Santos N, Gummadi K, Rodrigues R (2009) Towards trusted cloud computing. In: Proc of HotCloud [22] Krautheim FJ (2009) Private virtual infrastructure for cloud computing. In: Proc of HotCloud (http://static.usenix.org/event/hotcloud09/tech/full_papers/kraut heim.pdf) [23] Kailash Patidar, Ravindra Gupta, Gajendra Singh, Megha Jain, Priyanka Shrivastava, “Integrating the Trusted Computing Platform into the Security of Cloud Computing System”, International Journal of Advanced Research in Computer Science and Software Engineering, Volume 2, Issue 2, February 2012. [24] Sue-Chen Hsueh, Jing-Yan Lin, and Ming-Yen Lin, “Secure Cloud storage for convenient Data archive of smart phones”, 2011 IEEE 15th International Symposium on Consumer Electronics, pp. 156-161. [25] TCG Storage Architecture Core Specification, April 2009. [26] Zou P, Wang C, Liu Z, Bao D. “Phosphor: a cloud based drm scheme with SIM card”, In Proceedings of the 12th

[27]

[28]

[29]

[30]

International Asia-Pacific on Web Conference (APWEB), 2010; 459. S. Bouzefrane, A. Benkara Mostefa, F. Houacine, H. Cagnon, “Cloudlets Authentication in NFC-based Mobile Computing”, 2nd IEEE International Conference on Mobile Cloud Computing, Services, and Engineering, April 2014, pp.267 – 272, Oxford, U.K. D. Huang, X. Zhang, M. Kang, J. Luo, Mobicloud: building secure cloud framework for mobile computing and communication, in: Proceedings of the Fifth IEEE International Symposium on Service Oriented System Engineering, SOSE’2010, pp. 27–34. B.-G. Chun, S. Ihm, P. Maniatis, M. Naik, A. Patti, Clonecloud: elastic execution between mobile device and cloud, in: Proceedings of the Sixth Conference on Computer Systems, EuroSys’11, ACM, New York, NY, USA, 2011, pp. 301–314. Nicolae Paladi, “How to Secure Infrastructure Clouds with Trusted Computing Technologies”, Swedish Institute of Computer Science, April 2013.

Table 1. Comparison of features of trusted platforms Criteria

SE

TEE

TPM

HCE

Form

Comprised of software Made up of software and Hardware-Embedded in a Software-Android Applications and tamper resistant hardware host platform hardware Dependency Autonomous entity a secure area that resides in Bounded with the host User defined (I don’t know if it is platform user defined or OS Library (like a little PC) the main processor mobile defined) devices crypto-processor, Not available Hardware Processor, crypto- Processor, crypto-processor, EEPROM, RAM. The EEPROM,RAM processor, EEPROM, memories may be shared with RAM the host device OS Within the chip Separate and run in parallel Bootstrap in the TPM, Operate in the host OS with the host OS OS in the host platform API support

Limited

Limited

Security features

Stores sensitive data (PIN code, IMSI), secret keys, generate session keys

Provides a framework for security within the device

Limited

Rich API support

Trust root, generates Security is dependent on device OS asymmetric key pair, binding, identity attestation, protected objects Used to secure the host Used for the software emulation platform, to remote of a smart card-based application communication

Used for identification/ Used for robust, hardwareauthentication, for data backed, scalable-consistent, OS-independent security. encryption during communication No study No counter-measures for Physical Many studies (the chip Depending on hardware features of hosting platform physical attacks attacks is endowed with counter-measures) Software Many studies (the chip Resistant to software attacks Resistant to software low resistance to software attacks attacks is endowed with attacks counter-measures) Privacy Controlled by the user Controlled by the user No control by the user Controlled by the user Role