Product Authentication Service of Consumer's mobile RFID Device

Juhan Kim, Dooho Choi, Inseop Kim and Howon Kim

Abstract We propose a product authentication service which is anti-counterfeiting technology and can be applied to both the supply chain (SCM) and forthcoming mobile RFID environment. It aims to become aware of distribution of spurious products with fake RFID tags and to provide a product authentication service to consumers with mobile RFID devices like mobile phone which has a RFID reader. We also propose an authentication mechanism between a tag and an information server through the device. The mechanism is designed to be appropriated for EPC Class-1 Generation-2 UHF tags. The tags, one of RFID tag types, is expected that most companies are planning to use it in the SCM in the short term and in consumer packaging in the long term due to its inexpensive cost. Because of the very cost, however, its resources are extremely scarce and it is hard to have any valuable security algorithms in it. It causes security vulnerabilities, in particular cloning the tags for counterfeits. The mechanism provides secure mutual authentication between a tag and a server. It can be applied to the mobile RFID environment that any mobile RFID reader can be placed in the interface between the tag and the server. It means that the mechanism provides secure communication between them even the reader is not trusted. On basis of this tag authentication, the service makes use of location information especially from LBS(Location Based Service) of the consumers' devices. Thanks to it, the service can make fine-grained decision about authentication of a product. At the same time the server utilize the information to extend its management area where legacy SCM could not reach thanks to consumers' mobile RFID devices. 1 Index Terms Mobile RFID. Anti-counterfeiting, Tag authentication.

I. INTRODUCTION EPC Class-1 Generation-2 UHF (860-960 MHz) standard has recently ratified by EPCglobal which is the non-profit organization charged with promoting the adoption of Electronic Product Code (EPC) technology. This defines the physical and logical requirements for a passive-backscatter,
1 Juhan Kim, Dooho choi and Howon Kim are with Electronics and Telecommunications Research Institutes, Daejeon, Korea (e-mail: {juhankim, dhchoi, khw}@[Link]). And Howon Kim is a leader of RFID/USN Security Research Team. Inseop Kim is currently M.S candidate at University of Science and Technology, Daejeon, Korea (email: kis54388@[Link]).

Interrogator-talks-first (ITF), radio-frequency identification (RFID) system operating in the 860 MHz - 960 MHz frequency range. The system comprises Interrogators, also known as Readers, and Tags, also known as Labels. An Interrogator interacts with a Tag by modulating an RF signal in the 860 MHz - 960 MHz frequency range. The Tag receives both information and operating energy from this RF signal. Tags are passive, meaning that they receive all of their operating energy from the Interrogator's RF waveform [1], [2]. Thanks to relatively inexpensive cost to the tags, ideal estimate is 5 cents in next several years [1], [5], it is expected that most companies are planning to use it in the supply chain in the short term and in consumer packaging in the long term [20]. Because of the very cost, however, its resources are extremely scarce and it is hard to have any valuable security algorithms in it. It causes security vulnerability, in particular cloning the tags for counterfeits. Counterfeiting is one of the fastest growing economic crimes worldwide. It threatens the economies of developed and developing countries alike, undermines trading relations, scares off vital new investment, and increasingly endangers public health and safety. Counterfeiting has spread at an alarming rate to electrical and electronic goods and components, machines and equipment, spare parts of all types, pharmaceutical [10] and even high technology products. Counterfeit products account for between 5-7 percent of world trade. In this paper, we propose an application level solution against counterfeiting. It aims to become aware of distribution of spurious products with fake RFID tags and provide product an authentication service to general consumers with mobile RFID devices [18] like mobile phone or PDA which can connect mobile RFID readers [4], [16] which are made especially for the mobile phone and the PDA. Targeted RFID tags are compliant with EPC Class-1 Generation-2 UHF standard that is expected that most companies are planning to use it in the supply chain. We also propose a security mechanism that a tag and a server can authenticate mutually through a mobile reader, which has the product authentication service. The mechanism is conformed to the air interface of EPC Class-1 Generation-2 UHF tags. Without any interference of the interface, it can be applied to the tags. The mechanism can be adapted to the mobile RFID environment that any mobile RFID reader can be placed in the interface between the tag

1-4244-0216-6/06/$20.00 2006 IEEE

and the server. It means that the mechanism provides secure communication between them even the reader is not trusted. On basis of this tag authentication, the service makes use of location information especially from LBS(Location Based Service) of the consumers' devices. Thanks to it, the service can make fine-grained decision about authentication of a product. Our concerns are not only security techniques of tagreader level to prevent cloning tags but also tag-application level methods to provide a mobile service for consumers to check products' originality before purchasing them and for manufacturers to detect counterfeits by the mobile service. With the service and supply chain management system they have, manufacturers can find out where the counterfeits are and which distribution channels are in problem. As a result of the processes, it is possible to prevent mass distribution of the counterfeits. We present an overview of the EPC, EPCglobal network, mobile RFID environment and related works in section 2. In section 3, we propose an anti-counterfeiting solution and a security mechanism, and we conclude in section 4. II. BACKGROUND The EPC was created by the Auto-ID Center as an eventual successor to the bar code. The aim was to create a low-cost method of tracking goods using RFID technology. The benefit of RFID is that it doesn't require line-of-site, which means goods can be scanned through packaging and without needing people to scan items. EPC tags were designed to identify each item manufactured, as opposed to just the manufacturer and class of products, as bar codes do today [17]. A. EPC Network EPCglobal is a joint venture between EAN International and the Uniform Code Council (UCC). It is a not-for-profit organization entrusted by industry to establish and support the EPC Network as the global standard for immediate, automatic, and accurate identification of any item in the supply chain of any company, in any industry, anywhere in the world. Its objective is to drive global adoption of the EPCglobal Network. The EPCglobal EPC Network consists of five fundamental elements [17]: (1) EPC: The EPC is a globally unique serial number that identifies an item in the supply chain. This allows enquiries to be made about a single instance of an item, wherever it is within the supply chain. (2) The ID System: The ID System consists of EPC tags and EPC readers. EPC tags are RFID devices that consist of a microchip and an antenna attached to a substrate. The EPC is stored on this tag, which is applied to cases, pallets and/or items. EPC tags communicate their EPCs to EPC readers using Radio Frequency Identification. EPC readers

communicate with EPC tags via radio waves and deliver information to local business information systems using EPC Middleware. (3) EPC Middleware: EPC Middleware manages real-time read events and information, provides alerts, and manages the basic read information for communication to EPC Information Services as well as a company's other existing information systems. EPCglobal is developing a software interface standard for services enabling data exchange between an EPC reader or network of readers and information systems. (4) EPC-IS: EPC Information Services enables users to exchange EPC-related data with trading partners through the EPCglobal Network. (5) Discover Service: The Discovery Services is a suite of services that enables users to find data related to a specific EPC and to request access to that data. The Object Naming Service (ONS) is one component of the Discovery Services. B. Mobile RFID Environment In this paper, mobile RFID environment means that mobile RFID devices like mobile (cellular) phone and PDA employing cellular network can be used at querying information of the consumer product related a tag [12], [19] as shown in Figure 1.. The mobile RFID devices should have mobile RFID readers which are made for them and are, in particular, compliant with EPC Class-1 Generation-2 UHF (860-960 MHz) standard [18]. That is, Class-1 Generation-2 tags mainly for supply chain are also used for consumer products. Thus, any consumer can get information of a product with a tag by sending an EPC of the tag to an EPC information server with her mobile RFID device in this mobile RFID environment.
Root ONS

Mobile phone Product Authentication Service Local ONS LSB Provider

SCM

Mobile Network Middleware 1. RFID Reader 2. 3. 4. Mobile phone Local ONS : Query about EPC-IS address with EPC Code Local ONS< Root ONS : Query EPCIS address, and Get EPCIS address Local ONS Mobile : Provide EPCIS address Mobile>EPCIS : Query, and Get information related the EPC EPC-IS > EPC-PAS : Query about the product authentication, and get the result (Location Info. from LBS provider, ownership info. from SCM or phone) EPC-IS

yrarbiL ytiruceSF/I SICPEevloser SNOtnemeganaM redaeRtnemeganaM redaeRnoitcelloc & gniretliFRFID tag Product

EPC-PAS

5.

RFID Service Network

Fig 1. Overview of the mobile service for the product authentication based on the architecture of Mobile RFID Forum [12].

C. Related Works Legacy approaches like juels [5], [6], [8]. [9], molnar [7] and golle [11] are focused on making techniques which are impossible to clone tags by proving secure channel between tags and readers, or between tags and database. Those

approaches, however, are hard to be employed at EPC Class1 Generation-2 UHF (860-960 MHz) compliant tags because of limited resources the tags except juels [8]. Quiet briefly, juels makes use of kill PIN and access PIN for authentication by sending those PINs to a tag and analyzing responses from the tag. The kill PIN is a password for consumers' privacy by killing the tags operation completely and permanently and the access PIN is used to access data bank in the tag. The Juels's approach is very useful in supply chain. However, it is not suited for authenticating products by consumers' mobile RFID devices, because the PINSet for PIN-test for authentication are easily exposed to the consumer's mobile RFID device. Once any adversary with a mobile RFID device requests information of the product with a tag to a centralized server, she can get the PINSet with her mobile RFID device during the interaction of the tag and the server. Then she can kill the tag easily by sending PIN one by one in the PINSet. It is because the number of PINs in the PINSet is quite smaller than combination number of the kill-PIN consists of 32 bits. Staake [3] proposed EPC-PAS (Product Authentication Service) for anti-counterfeiting which can track and trace the movement of goods from production to consumption and which employs secure authentication in a Database-ReaderTag environment like Figure 2. The secure authentication, however, cannot be applied to the tags compliant with the standard.
Root ONS Request/Respons Address e Update

recently a mobile RFID reader for PDA like IP4 [18] is launched. In this paper, we assume that a mobile RFID device is kind of mobile phone or PDA that can be connected to CDMA or GSM network and has such a RFID reader. A. Mobile RFID Environment Thanks to above devices, it is possible to track and trace a product during whole life-cycle of typical RFID tag that consists of four main phases such as supply chain, point-ofsales, customer control & after sales service, and recycling & waste management. In the mobile RFID environment that the mobile RFID devices can be used at, the EPC-PAS can track and trace a product with a tag whenever a consumer requests information of the product with her mobile device like Figure 3.
Root ONS Request/Response Address Update Mobile Access

Local ONS RFID Service Network

Local ONS

Mobile Network

EPC-IS

EPC-PAS

EPC-DS EPC-IS

Mobile RFID Device

Product with EPC and watermark Manufacturer Distributor Anywhere Consumer

Fig 3. The EPC infrastructure we are evolving for anti-counterfeiting which is irrespective of the location of a product with a tag.

Local ONS

3 EPC-IS EPC-DS EPC-PAS

EPCIS Product with EPC Retailer

Manufacture r

Distributor

Fig 2. Staake's evolving EPCglobal network including the EPC-PAS for anti-counterfeiting.

In this respect, it is expected that manufactures can get a powerful means to prevent mass distribution of counterfeits with only extending ability about tracking and tracing to the mobile RFID environment. Therefore, even if there are counterfeits at markets which is not covered with the supply chain, manufacturers can find out them and which distribution channels are in problem as soon as any consumer requests information of the product with a tag. As a consequence, it is possible to prevent mass distribution of the counterfeits and improve the distribution channel for anti-counterfeiting. B. The EPC-PAS In proposed works, the EPC-IS supports an interface for mobile RFID devices. The EPC-IS that receives a request from the consumer invokes the EPC-PAS for product authentication in like Figure 4. Then the EPC-IS gets the result from it and sends information of the product with the result to the consumer and it accumulates the location to build historical locations for tracing a product. The EPC-PAS has an LBS client for taking location information from LBS providers and

III. PROPOSED ANTI-COUNTERFEITING SOLUTION The reason that we employ Class-1 Generation-2 standard is because it is expected that most companies are planning to use it in the supply chain and in packaging consumer products at the same time. Mobile RFID readers for consumers, which are used to read such consumer products that are compliant with the standard, are now developing and

provides alert service to warn manufacturers of discovering the counterfeits.

Manufacturers Application A consumer requests authentication service about a product Product Certificate EPC RFID tag Alert service Product Authentication Service

Loc. Info. LBS Client Watermark

Mobile device with mobile RFID reader

EPC Loc. Info. Owner Info.

Matching Module

Extracting watermark Context History Loc. owner

Product

EPC Tracking Database

Loc. Info.

EPC-IS EPC-PAS
EPC Product Certificate

Context History Loc. owner

EPC

Manufacturer

Distributor

Retailer

Consumer request authentication service about products

Fig 4. The dataflow of the EPC-PAS we design, which has a product authentication service for consumers and an alert service for manufacturers.

The EPC-PAS accumulates contexts like location information from LBS providers and ownership information from consumers' devices or supply chain management systems whenever the consumers query about information of a product. Then it returns a certificate which comprise of the authentication result like 'a real article', 'not real article', or 'can not authenticate', date of issue, date of expiry, destination information, product information and PKI-based digital signature [21] of it. Mobile RFID devices for consumers, therefore, should support functions that can identify manufacturer and check integrity of the certificate by verifying the signature. It should also check validation date in the certificate. C. Tag Authentication Scheme In our proposed scheme, we only concentrate on identifying Tag to Verifier. Therefore, our protocol does not support the un-traceability of RFID tag. We assume that the following functions are implemented into RFID tag: F, G : {0,1}l X {0,1} l {0,1}l are one-way functions R is an l-bit random number generator In our scheme, Tag and Verifier share the initial secret keys, T0, T0', T0'', S0, S0', S0'' and then these keys will be updated for each authentication round. The protocol consists of Initial Round, Key Update, n-th Round, and Key Refreshment. Details of our scheme are as follows:

In the n-th round, Tag does not respond to An, Bn and the authentication process is halted, if An Bn Sn S n 0 at the step 2.

At the step 3, Verifier returns failure and should refresh his keys with his current keys, if the messages Cn, Dn, En are not delivered at Verifier or Cn Dn En Tn Tn 0. Otherwise, Verifier responds with Fn, updates keys for the next round, and stores the current keys Tn,Tn,T'n,Sn,S'n,S''n. Tag updates his keys and stores his current keys if Fn S''n R''n = 0. Otherwise, he does not respond to any authentication message and should refresh his keys with his stored last keys.

Then the adversary A 's advantage Adv(A) is defined to be Adv(A) = Pr[ (T0, T0', T0'', S0, S0', S0'')] Setup ; R (C*, D*, E*) O : V (C*, D*, E*) = (accept, F*) ] A Here, O is defined as follows: In Test Phase, O = { T, V } In Cloning Phase, O = {V } And Setup is a process that T and V share the initial secret keys. In our security model, the tag and verifier oracles may invoke Key Refreshment at any time. Definition 1: For a given adversary A, RFID authentication scheme is said to be (t, qT, qV, )-secure if A runs in time at most t, makes at most qT queries to the tag oracle T and at most qV queries to the verifier oracle V, and Adv(A) . Theorem 2: Suppose that F and G are random oracles. Then the proposed RFID authentication scheme is (t, qT, qV, )-secure, where = (qT + qV + 1) . Proof: Let A be an adversary and G0 be an original attack game. Then A produces a triple (C*,D*,E*), even after obtaining qT tag's response messages and qV verifier's messages of its choice from the tag oracle and the verifier oracle respectively. Let S0 be the event that V (C*,D*,E*) = (accept, F*). Then Adv(A) = Pr[S0]. Now we will define modified attack games G1 and G2. Let Si be the event that V (C*,D*,E*) = (accept, F*) in Gi for i=1,2. Game G1. Let A1 be a player of G1. In G1, A1 modifies the game G0. Suppose A makes the tag oracle query for (A,B)i, 1 i qT . If (A,B)i is the original output of the verifier oracle V, A1 returns the tag oracle query result for (A,B)i. Otherwise, A1 returns . Let W be an event of G0 such that (A,B)i is not the original output of V, but T returns valid response for (A,B)i. Then Pr[W] qT , since the secret keys
2 2l

We suppose that all messages of the key refreshment process are delivered via a secure channel. At the step 1, Verifier extracts his new keys T0,T'0,T''0 from Tn || T'n || T''n T0 || T'0 || T''0. At the step 2, Tag also extracts his new keys S0,S'0,S''0 from Sn || S'n || S''n S0 || S'0 || S''0. Tag and Verifier use the exchanged T0,T'0,T''0,S0,S'0,S''0 as their new keys. D. Security Analysis In this section, we describe the security model of our scheme and prove the formal security of the proposed scheme. The security model is similar to the authentication security definition of [14] and the security definition of digital signature. At [14], Ari Juels defined the notion of authentication security for the tag authentication protocol. The definition of authentication security for the scheme characterizes the ability of an adversary A to forge a valid tag's response message so that the verifier oracle V returns accept at the last step of the authentication process. Before the adversary A forges a valid tag's message, he can query the messages with a tag oracle T and the verifier oracle A. Therefore, our attack scenario has two phases as follows: Test Phase: A interacts with the tag oracle T and the verifier oracle V and so, he collects the query results. Cloning Phase: A queries V until A yields some output, either (accept, F*) or , where is empty.

are one-time passwords in our scheme. Therefore, Pr[S0] qT Pr[S1]. Pr[W] + Pr[S1]
2 2l

Game G2. Let A2 be a player of G2. In G2, A2 modifies the game G1. The verifier oracle query results have two types as follows: (Type I) The query result for the tag oracle identification, ID. It is a challenge message of the authentication scheme. (Type II) The query result for (C,D,E), where (C,D,E) is the tag response message. It is a message of the second step of the authentication scheme. If a query result is the type I query, A2 returns the challenge message of the verifier oracle for ID. Suppose x is the type II query. Then A2 returns the message of the verifier oracle for x, if x belongs to the query results of the tag oracle and A2 returns otherwise. Let U be an event of G1 such that x does not belong to the query results of the tag oracle,

2 2l

but the verifier oracle returns (accept, F*). Then Pr[U] qV . Therefore, Pr[S1] Pr[U] + Pr[S2] qV +Pr[S2].
2 2l 2 2l

Basically, A2 is just an eavesdropper of the real world. Therefore, Pr[S2] 1 , since F and G are random oracles
2 2l

and our scheme is based on one-time padding. Hence, Pr[S0] Pr[U] + Pr[S2] qV + 1 = (qV + 1) .
2 2l 2 2l 2 2l

From the above results, Adv(A) qT + (qV + 1) = (q T + qV + 1) 2l 2l 2l

2 2 2

IV. CONCLUSION We proposed an application level anti-counterfeiting solution that can track and trace a product through whole life-cycle of EPC tag with upcoming consumers' mobile RFID devices like mobile phone or PDA with a mobile RFID reader. We also suggested tag authentication scheme which is tag-level security mechanism and which is required to the solution. We extended the EPC-PAS to cover mobile RFID environment for tracking and tracing a product continuously with an EPC tag, irrelative of places, whenever a consumer requests information of the product with her mobile RFID device. Therefore, even if there are counterfeits at market, manufacturers can find out them and which distribution channels are in problem as soon as any consumer requests information of a counterfeit with a fake tag. REFERENCES
[1] [2] EPCglobal, [Link] EPCglobal, EPC Radio-Frequency Identity Protocols Class-1 Generation-2 UHF RFID Protocol for Communications at 860 MHz 960 MHz Version 1.0.9, [Link] Thorsten Staake et al., Extending the EPC Network - The Potential of RFID in Anti-Counterfeiting, ACM symposium on Applied computing, 2005. Nokia unveils RFID phone reader, RFID Journal, March 2004. Ari Juels, Minimalist cryptography for low-cost RFID tags, SCN 04, 2004. Ari Juels, Yoking-proofs' for RFID tags, In PerCom Workshops 2004. David Molnar et al., Privacy and Security in Library RFID : Issues, Practices, and Architectures, In Computer and Communications Security, 2004. Ari Juels, Strengthening EPC Tags Against Cloning, ACM Workshop on Wireless Security(WiSe), 2005. Ari Juels, Squealing euros: Privacy protection in RFID enabled banknotes, In FC'03, January 2003. U. S. Department of Health and Human Services, Food and Drug Administration (2004) Combating Counterfeit Drugs, [Link] P. Golle, M. Jakobsson, A. Juels, P. Syverson, Universal reencryption for mixnets, CT-RSA'04, 2004. Mobile RFID Forum, [Link] RFID Journal, Frequently Asked Questions, [Link]

[14] Ari Juels, Minimalist Cryptography for Low-Cost RFID Tags, The Fourth International Conference on Security in Communication Networks . SCN 2004, LNCS, Springer, 2004. [15] RFID Journal, The 5-Cent Challenge, [Link] August 2004. [16] RFID Journal, Test Set for RFID-Enabled Phones, [Link] September 2004. [17] EPCglobal, About the EPCglobal Network, [Link] [18] Intermec, P4 Portable RFID Reader, [Link] eprise/main/Intermec/Content/Products/Products\\\_ShowDetail?Prod uct=RFID2\_IP4. [19] IT Korea Journal, Mobile RFID Forum' Launched, March~April 2005, [Link] [Link] [20] ICC Counterfeiting Intelligence Bureau, The International AntiCounterfeiting Directory 2003, [Link] _bureau/[Link]. [21] RSA Laboratories, What is the RSA cryptosystem?, [Link] [22] Stephan J. Engberg et al., Zero-knowledge Device Authentication: Privacy & Security Enhanced RFID preserving Business Value and Consumer Convenience, PST, 2004.

Juhan Kim received his B.S degree from ChungNam National University, Daejeon, Korea, in 1997 and the M.S degrees in Computer Science from the same university 1999. He is currently a senior member of engineering staff at the Electronics and Telecommunications Research Institute (ETRI), Daejeon, Korea. Currently, his main research is focused on RFID and WSN security issues. Dooho Choi received his B.S degree from SungKyunGuan University, Seoul, Korea, in 1994 and the M.S and Ph.D. degrees in Mathematics from Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea, in 1996 and 2002, respectively. He is currently a senior member of engineering staff at the Electronics and Telecommunications Research Institute (ETRI), Daejeon, Korea. His current research projects include security for the RFID & USN. Inseop Kim received the B.S degree from Dankook University in 2005. He is currently a M.S candidate at university of science and Technology (UST) and research student at ETRI. His current research projects include security for the RFID & USN.

[3]

[4] [5] [6] [7]

[8] [9] [10]

[11] [12] [13]

Howon Kim received his B.S.E.E. degree from KyungPook National University, DaeGu, Korea, in 1993 and the M.S and Ph.D. degrees in Electronic and Electrical Engineering from Pohang University of Science and Technology (POSTECH), Pohang, Korea, in 1995 and 1999, respectively. From July 2003 to June 2004, he studied at the COSY group at the RuhrUniversity of Bochum, Germany. He is currently a team leader at the Electronics and Telecommunications Research Institute (ETRI), Daejeon, Korea. His research interests include information security and computer architecture. Currently, his main research focus is on private and public key crypto processor design and RFID, WSN security issues. He is a member of the IEEE, IEEE Computer Society, and IACR.