FIELD
The present invention relates generally to Radio Frequency IDentification (RFID). More particularly, the present invention relates to secure two-way RFID communications.
BACKGROUND
Radio Frequency IDentification (RFID) systems are used for identifying and tracking items, inventory control, supply chain management, anti-theft of merchandise in stores, and other applications. As shown in FIG. 1, a typical RFID system 10 consists of a plurality of transponders (referred to in the art as “tags”) 100-0, 100-1, . . . ,100-N and one or more transceivers (referred to in the art as a “readers”) 102. A reader 102 includes an antenna 104, which allows it to interrogate one or more of the tags 100-0, 100-1, . . . ,100-N over a wireless link 106. The tags 100-0, 100-1, . . . ,100-N also have their own respective antennas 108-0, 108-1, . . . ,108-N, which allow them to transmit tag information back to the reader 102 over reverse links 107-0, 107-1, . . . ,107-N. The reader 102 may then use this tag information as a look-up key into a back-end database 110, which stores product information, tracking logs, key management data, and the like.
In order for the reader 102 to address any particular tag from the population of tags 100-0, 100-1, . . . ,100-N, a process known as “singulation” is commonly used. To singulate a tag from the population of tags 100-0, 100- 1, . . . ,100-N, the reader 102 polls the tags 100-0, 100-1, . . . ,100-N for their ID numbers. Because multiple tag responses may interfere with one another, anti-collision algorithms are typically employed in the singulation process. Anti-collision algorithms are either probabilistic or deterministic. One well-known probabilistic anti-collision algorithm is the Aloha technique, whereby tags 100-0, 100-1, . . . ,100-N respond to a polling signal from the reader 102 at random intervals. If a collision occurs, the tags responsible for the collision wait for another, usually longer, time interval before responding again. A known deterministic anti-collision algorithm is the so-called “binary tree-walking” algorithm. According to this approach, the reader 102 initially polls the tags 100-0, 100-1, . . . ,100-N for the first bit of the tags' respective ID numbers. Based on the bit values received, the reader 102 then limits the number of tags which are to send subsequent bits of their ID numbers. This process is repeated until the ID of a single tag has been singulated.
A tag is usually embodied as a semiconductor microchip having a small amount of memory for storing the tag's ID number and, in some applications, information concerning the item to which the tag is associated. Further, tags are either “passive” or “active”, depending on how they are powered. An active tag contains its own on-board power source, i.e., a battery, which the tag uses to process received signals and to transmit tag information back to a reader. A passive tag does not have its own on-board power source. Rather, it derives the power it needs by extracting energy from the RF carrier signals broadcast by the reader. The passive tag transmits information to the reader using a process known as modulated backscattering, a process which is described in more detail below. Because passive tags do not have their own power sources, and rely on backscattering, they cannot be read from great distances. Nevertheless, they have, in many applications, become more popular than active tags since they are less expensive to manufacture, maintain, and operate.
In a conventional passive-tag-based RFID system, a tag derives its power from a CW (continuous wave) RF (radio frequency) carrier signal sent from a reader over a forward link 204. As shown in FIG. 2, a tag 200 also modulates the CW signal using modulated backscattering, a process by which the antenna matching network impedance is varied depending on the information being provided by the tag. For digital information, the antenna terminal may be simply switched by the tag's modulating signal, from being an absorber of RF radiation to being a reflector of RF radiation. In this manner the tag's information is encoded on the CW signal and backscattered back to the reader 202 over a reverse (or “backscatter” link) 206.
Whereas RFBD systems provide a useful system for identifying and tracking objects, such systems are subject to a number of privacy and security risks. These security risks can arise during polling, singulation, and following singulation when a reader is communicating one-on-one with a particular tag. Without adequate access control, unauthorized (i.e., “rogue”) readers may be able to interrogate tags or intercept information, which would otherwise remain secret. (FIG. 2 shows, for example, an eavesdropper 208 intercepting a backscattered signal from the tag 200.) Further, rogue (or “spoofed”) tags, which have been made or modified to appear as authentic tags, may be able to gather information from legitimate readers.
In addition to the security concerns just described, RFID systems without proper security and privacy measures in place undesirably allow unauthorized “location tracking”. Unauthorized location tracking allows one or more readers to track RFID-labeled items (e.g., clothing worn by an individual or items an individual may be carrying such as tagged smart cards, credit cards, banknotes, and the like). Consequeritly, without proper access control or prevention measures in place, the privacy normally taken for granted concerning an individual's movement, social interactions and financial dealings can be compromised by RFID systems.
Various proposals for addressing the security and privacy risks associated with RFID systems have been proposed. One technique that has been proposed to avoid unauthorized access to readers and tags of an RFID system is “symmetric encryption”. According to this technique, special encryption and decryption hardware is built into both the readers and the tags of the RFID system. A block diagram of a symmetric encryption RFID system is shown in FIG. 3. A drawback of the symmetric encryption approach, however, is that a large number of logic gates (e.g., between 20,000 and 30,000) is required to implement the encryption and decryption hardware. This increases the size and complexity of the microchip embodying the tag. Consequently, symmetric encryption is not a technique that readily allows the manufacture of small and inexpensive tags. For at least this reason, therefore, symmetric encryption is not a favorable solution to RFID.
Another technique that has been applied to avoid the security and privacy concerns described above is a technique known as “public-key” encryption. Use of public-key encryption permits a tag to transmit encrypted information, together with a public key known by both the reader and the tag, to the reader. The reader, having a private key known only to it, is then able to decrypt the information communicated by the tag. Unfortunately, similar to the symmetric encryption approach, public-key encryption requires a large number of logic gates (e.g., more than 30,000 logic gates) to implement the encryption hardware. Accordingly, for reasons similar to those associated with use of symmetric encryption, public-key encryption is not a simple and cost-effective solution to RFID.
Whereas many existing and proposed RFID systems prove to be prohibitively expensive for widespread deployment, others make assumptions that, if built into an RFID system, do not sufficiently respect the security and privacy concerns discussed above. An example of such a security and privacy compromised RFID system is described in “Security and Privacy Aspects of Low-Cost Radio Frequency Identification Systems,” by Stephen A. Weis, Sanjay E. Sarma, Ronald L. Rivest and Daniel W. Engels, First International Conference on Security in Pervasive Computing (Mar. 12-14, 2003). The RFID systems proposed in that paper assume that it is only possible for an eavesdropper to monitor the forward link (i.e., signals sent from the reader to the tags). In other words, it is assumed that the power in the link from the tag to the reader (i.e., the backscatter link) is so weak, and/or that any possible eavesdropper is at such a large distance away from the tag, that an eavesdropper could not possibly intercept information from it. It also makes the assumption that security can be enhanced, simply by reducing the power in the backscatter link. For a number of reasons described below, however, an RFID system designed using these assumptions would have reduced security and privacy effectiveness.
First, because tags of a passive-tag RFID system extract their power from the carrier on the forward link (i.e., the reader-to-tag link), the power of the signal in the forward link must be large enough so that sufficient power is available for the tag to operate. This means that the power in the backscatter link can be quite large. Accordingly, the assumption that the power in the backscatter link is so weak that an eavesdropper cannot intercept it is not necessarily a fair assumption. Second, even if it is assumed that an eavesdropper is a large distance away from the tag, this large distance may, in many circumstances, be overcome simply by using a larger eavesdropper antenna. Finally, even if power in the backscatter link could be reduced by lowering the power in the forward link to enhance security, not only would the range of the RFID system be limited and consequently have diminished utility, such an approach could also be defeated, again simply by using a larger eavesdropper antenna.
SUMMARY
Methods and apparatuses for providing secure two-way (reader-to-tag and tag-to-reader) RFID communications are disclosed. According to one aspect, an RFID reader includes a signal generator that is adapted to generate an RF carrier signal and modulate it to noise encrypt the RF carrier signal, which can include any signal(s) not known to an unintended or unauthorized recipient (i.e., an unintended or unauthorized reader, tag, or eavesdropper). A tag receives the noise-encrypted RF carrier signal and backscatter modulates it with tag information. The tag information may comprise the tag's ID number or other information associated with the item to which the tag is attached. Eavesdroppers cannot extract the tag information from the backscattered signal because it is masked by the noise encryption.
Other aspects of the inventions are described and claimed below, and a further understanding of the nature and advantages of the inventions may be realized by reference to the remaining portions of the specification and the attached drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
The accompanying drawings, which are incorporated into and constitute a part of this specification, illustrate one or more embodiments of the present invention and, together with the detailed description, serve to explain the principles and implementations of the invention.
In the drawings:
FIG. 1 shows a typical prior art RFID system.
FIG. 2 shows a prior art passive-tag RFID system, illustrating the forward link with its continuous wave (CW) signal, the reverse (or “backscatter” link), and an eavesdropper intercepting a backscattered signal.
FIG. 3 shows a prior art symmetric encryption RFID system, highlighting the fact that both the tag and reader include substantial hardware components.
FIG. 4 shows an RFID system, according to an embodiment of the present invention.
FIG. 5 shows the backscattered frequency domain baseband equivalent spectrum of a backscattered signal, in which no amplitude or phase modulation has been applied to the reader carrier signal, as might be found in the prior art.
FIG. 6 shows the backscattered frequency domain baseband equivalent spectrum of a noise modulated (i.e., A(t)≠1 and θ(t)≠0) backscattered signal, according to an embodiment of the present invention.
FIG. 7 shows a waveform that might be used to encrypt a reader RF carrier signal instead of a CW waveform.
FIG. 8 shows a waveform of a signal backscattered from a tag in response to encryption by the waveform of FIG. 7.
FIG. 9 shows an RFID system, which applies AM noise to the reader carrier signal, according to an embodiment of the present invention.
FIG. 10 shows an RFID system, which applies FM/PM to the reader carrier, according to an embodiment of the present invention.
FIG. 11 shows a timing diagram illustrating a method of establishing a secure two-way communication link between a reader and a tag of a population of tags, according to an embodiment of the present invention.
FIG. 12 shows a timing diagram illustrating a method of establishing a secure two-way communication link between a reader and a tag of a population of tags, including applying a password lock to a singulated tag, according to an embodiment of the present invention.
FIG. 13 shows how, in establishing a secure two-way communication link according to embodiments of the present invention, a rogue reader is prevented access to information backscattered by a tag.
FIG. 14 shows how, in establishing a secure two-way communication link according to embodiments of the present invention, a rogue tag is prevented from communicating with a legitimate reader.
FIG. 15 shows an analog implementation of an RFID system, according to an embodiment of the present invention, in which both AM and FM/PM are used to modulate an RF carrier signal.
FIG. 16 shows an analog implementation of an RFID system, in which AM is used to modulate the carrier signal, according to an embodiment of the present invention.
FIG. 17 shows an analog implementation of an RFID system, in which FMIPM is used to modulate the carrier signal, according to an embodiment of the present invention.
FIG. 18 shows a combined analog and digital implementation of an RFID system, in which both AM and FM/PM are used to modulate an RF carrier signal, according to an embodiment of the present invention.
FIG. 19 shows a combined analog and digital implementation of an RFID system, in which AM is used to modulate an RF carrier signal, according to an embodiment of the present invention.
FIG. 20 shows a combined analog and digital implementation of an RFID system, in which FM/PM is used to modulate an RF carrier signal, according to an embodiment of the present invention.
FIG. 21 shows a digital implementation of an RFID system, according to an embodiment of the present invention.
DETAILED DESCRIPTION
Embodiments of the present invention are described herein in the context of methods and apparatuses relating to secure two-way RFID communications. Those of ordinary skill in the art will realize that the following detailed description of the present invention is illustrative only and is not intended to be in any way limiting. Other embodiments of the present invention will readily suggest themselves to such skilled persons having the benefit of this disclosure.
Reference will now be made in detail to implementations of the present invention as illustrated in the accompanying drawings. The same reference indicators will be used throughout the drawings and the following detailed description to refer to the same or similar parts.
Referring first to FIG. 4, there is shown an RFID system 40, according to an embodiment of the present invention. RFID system 40 comprises a reader 402 and one or more tags 400. Although not shown in FIG. 4 or other drawings in the disclosure, those skilled in the art will readily understand that both the reader 402 and tags 400 have antennas that permit the reader 402 to communicate with the tags 400 over an RF forward link 404 and the tags 400 to receive and backscatter RF signals back to the reader 402 over an RF backscatter link 406.
To communicate with a tag 400, the reader 402 broadcasts an RF signal to the tag 400. The RF signal is a continuous wave carrier signal, cos(ωt). Further, encryption is applied. The encryption may be implemented by modulating the RF signal with an amplitude modulation signal, A(t). The encryption may be further enhanced by adding a phase modulation signal, θ(t) to the AM modulated encrypted RF signal. For purposes of this disclosure, θ(t) represents either or both frequency modulation (FM) and phase modulation (PM). Accordingly, at various instances throughout the disclosure, the notation “FM/PM” will be used to indicate that either or both phase modulation and frequency modulation may be used to establish θ(t). The amplitude and phase modulated carrier signal is shown in FIG. 4 as A(t)cos(ωt+θ(t)). The amplitude modulation, A(t), and phase modulation, θ(t), are only known by the reader 402. Accordingly, together they serve as an encryption key. Note that if no encryption were present in the forward link signal, A(t) would equal unity and θ(t) would equal zero.
Upon receipt of the A(t)cos(ωt+θ(t)) signal by the tag 400, the tag 400 extracts power from the RF energy in the signal. The tag 400 also backscatter modulates A(t)cos(cωt+θ(t)) with a tag modulation signal (1+m(t)). The tag modulation signal (1+m(t)) contains identification information associated with tag 400, e.g., the tag's ID and/or information concerning the item to which the tag is associated. This information becomes masked by the amplitude and phase modulation noise provided by the A(t)cos(ωt+θ(t)) signal during backscattering, thereby providing an encrypted backscattered signal.
The reader 402 receives the backscatter modulated signal and amplifies it, for example by way of an automatic gain control (AGC) amplifier, sufficiently enough so that the reader receiver hardware is able to operate in the proper range. nR(t) in the drawing represents thermal noise that is unavoidably added to the received signal. Since the reader knows A(t) and θ(t), their inverses can be mixed with the received signal to remove the encryption caused by A(t) and θ(t). The resulting signal is then low-pass-filtered to remove the double frequency products generated by the mixer and other high frequency noise. The result at the output of the LPF (low pass filter) is the desired baseband signal, i.e., (1+m(t)), plus some unavoidable noise component, n1(t).
Also shown in FIG. 4 is an eavesdropper 408. The eavesdropper 408 is not part of the system 40, but is shown in FIG. 4 to illustrate how it might attempt to intercept transmission of backscattered signals in the backscatter link 406. If the eavesdropper 408 is somehow in range to receive the backscattered signal, it would have to first perform some AGC action to amplify the received signal, similar to what the reader 402 does. The frequency spectrum of the received signal would be similar to what the reader 402 receives. However, unlike the reader 402, the eavesdropper 408 has no knowledge as to what the amplitude modulation signal, A(t), looks like or whatθ(t) is. Consequently, the eavesdropper 408 can only mix with a local oscillator that does not have any information relating to the inverses of A(t) or θ(t).
The eavesdropper 408 might contain a phase locked loop (PLL) and a mixer, followed by an LPF, to produce a baseband signal. Alternatively, an envelope detector might be used, if the FM/PM in the received signal cannot be tracked using a PLL (phase locked loop). Use of an envelope detector would introduce additional degradations to the signal (i.e., in addition to the noise masking effect caused by A(t) and θ(t)), which would further reduce the likelihood that the eavesdropper 408 could ever succeed at actually extracting tag information from the backscattered signal. Assuming that either a PLL/mixer and LPF or an envelope detector are used, the LPF would also have to have a much higher cutoff frequency than the LPF used by the reader 408. The reason for this is that, because the eavesdropper 408 cannot remove the AM and possibly the FM/PM components at the front-end, the tag information signal (1+m(t)) remains spread over a broader frequency range than the “de-spread” signal produced by the reader 402. Consequently, the eavesdropper 408 would require the use of an LPF having a much greater cutoff frequency than that of the LPF used by the reader 402. The required use of a broader band LPF presents additional problems to the eavesdropper 408, since additional noise not filtered by the LPF, and introduced in the baseband signal, further decreases the likelihood that the eavesdropper 408 could ever determine the tag information signal (1+m(t)).
Even if the eavesdropper 408 was somehow successful at removing the FM/PM component, there would still remain the AM component, which masks the tag information signal (1+m(t)). At best, all the eavesdropper could ever obtain at baseband is the baseband signal, A(t)(1+m(t))+n2(t), i.e., the product of two time varying functions and a noise component, n2(t). The eavesdropper 408 does not have knowledge of A(t) or (1+m(t)) separately. Consequently, the backscattered signal cannot be decrypted by the eavesdropper 408, and the information in the tag information signal (1+m(t)) cannot be ascertained by the eavesdropper 408.
The noise masking effect caused by amplitude modulating and phase modulating the reader interrogation carrier signal can be seen by comparing FIG. 5 to FIG. 6. FIG. 5 shows the backscattered frequency domain baseband equivalent spectrum of a backscattered signal in which no amplitude or phase modulation has been applied to the reader carrier signal (i.e., where A(t)=1 and θ(t)=0). Distinct peaks (i.e., 500, 510, 520, . . . and 510′, 520′, 530′, . . . ) corresponding to bits of information in the tag modulation signal (1+m(t)), can be seen. This is an unfavorable situation, as it raises the possibility that the bits of information can be intercepted by a rogue reader. FIG. 6, by comparison, shows the backscattered frequency domain baseband equivalent spectrum of a noise modulated (i.e., A(t)≠1 and θ(t)≠0) backscattered signal, according to an embodiment of the present invention. As can be seen, the noise fills up the channel and masks (i.e., it covers up) the spectral shape of the tag modulation signal (1+m(t)).
The noise masking effect can be further seen by comparing the baseband waveforms of the reader 402 and an eavesdropper 408 in the time domain. FIG. 7 shows an encrypted RF carrier signal, which can be transmitted for encryption instead of a continuous wave (CW) waveform. Encryption is accomplished by modulating the reader's RF carrier signal. This modulation can be accomplished by amplitude modulation (AM) and can be further enhanced by PM/FM. PM/FM modulation alone, however, without the AM modulation will not achieve the desired high level of encryption. This can be understood as follows. The waveform of FIG. 7 is, in effect, the signal waveform (modulated bit stream encoded in a convenient manner and optionally digitally encrypted) that is to be received from the tag's backscattered signal after noise decryption (removal of the noise attributable to A(t) and θ(t)). On the other hand, FIG. 8 shows such a baseband waveform of a backscattered signal where the noise attributable to A(t) and θ(t) has not been properly removed, as might be received by an eavesdropper lacking knowledge of the noise sequences responsible for A(t) and θ(t). As can be seen from FIG. 8, the amplitude of the “bits” sequentially embedded in the signal varies wildly and bit values therefore cannot be accurately discerned. Consequently, from the eavesdropper's perspective it is difficult, if not impossible, to determine whether any given bit in the stream is a one or a zero. In the case of a recovered signal as illustrated in FIG. 7, however, the legitimate reader can and has mixed in the inverted A(t) and θ(t) and thereby removed the noise attributable to A(t) and θ(t) since it knows apriori the noise sequences that produce A(t) and θ(t) noise modulation components.
While the RFID system shown in FIG. 4 modulates the reader carrier signal using both AM and FM/PM, alternative embodiments could use one or the other. Accordingly, FIG. 9 shows an RFID system, which applies AM to the reader carrier, according to an embodiment of the present invention. Because only the reader has knowledge of the characteristics of the AM applied, an eavesdropper cannot decrypt tag information backscattered from a tag.
FIG. 10 shows an RFID system, which applies FM/PM to the reader carrier, according to an embodiment of the present invention, in addition to amplitude modulation. Because only the reader has knowledge of the characteristics of the FM/PM applied, an eavesdropper cannot decrypt tag information backscattered from a tag.
Referring now to FIG. 11, there is shown a timing diagram illustrating a method of establishing a secure two-way communication link between a reader and a tag of a population of tags, according to an embodiment of the present invention. According to this method, secure links are established both in the reader-to-tag direction and in the tag-to-reader direction. Because the method maintains two-way security during the entire time the secure two-way communication link is being established, rogue readers and rogue tags are prevented from intercepting and deciphering communications. Further aspects of the method, described in detail below, also prevent location tracking.
At step 1100 in the method shown in FIG. 11, a reader initiates communication by polling a population of tags, e.g., by broadcasting a polling signal having a random or pseudorandom ID. In response to the polling signal, the tags backscatter one or more bits. According to one embodiment, the backscattered bits from each tag are bits of pseudorandom numbers generated by a pseudorandom number (PN) generator on the tags. Using a tree-walking scheme, the reader responds, for example, by communicating that it only wishes to communicate with, for example, tags that transmitted bits of logic value “1”. Because the tags respond to each polling signal with one or more bits of a pseudorandom number, eventually a single tag is singulated. Whereas a binary tree-walking scheme has been described, those skilled in the art will readily understand that other singulation and anti-collision algorithms (probabilistic or deterministic) may be used to singulate the tag. Further, whereas singulating a tag has been described by use of a PN generator on the tag, singulation may be performed by simply using unique information stored on the tag (i.e., irrespective of whether a PN generator is on the tag).
Next, at step 1102, the singulated tag backscatters back to the reader a partial key, H(N), and a one-time pad pseudorandom number, R1-time pad. The one-time pad, R1-time pad, may have a value that is time independent or may have a value that may be changed over time. Further, it may be generated by the tag or simply stored on (but not necessarily generated by) the tag. Whereas both the partial key, H(N), and one-time pad are used in step 1102, in alternative embodiments of the invention either of the partial key, H(N), or one-time pad, R1-time pad, alone may be used. Noise encryption, as for example described above in relation to FIGS. 4-10, and denoted by “RE” in FIG. 11, is used to further encrypt the backscattered signal in this step 1102.
Upon receipt of the backscattered signal, at step 1104 the reader consults a secure back-end database to determine whether the value of H(N) sent from the tag is valid and, accordingly, whether the tag is authentic. If the reader determines that H(N) is a valid partial key, the method continues to step 1106. Otherwise, the reader discontinues communications with the tag, assuming that it is not authentic.
If the reader verifies that the tag is authentic, at step 1106 the reader transmits the other portion of the key, N, on the forward link to the tag. According to one embodiment, N is encrypted with a finction that depends on a pseudorandom number, which may be, for example, the one-time pad, R1-time pad, which was backscattered by the tag in step 1102. In FIG. 11, the encryption is shown as Nˆf(R1-time pad), the “ˆ” symbol indicating an exclusive OR (XOR) logic operation. Those skilled in the art will readily understand that an XOR operation is not required to form the encrypted key, and that other encryption schemes may be employed. The XOR operation is used in the described exemplary embodiment as it is computationally inexpensive.
Next, at step 1108 the tag verifies the authenticity of the reader, based on the value of the partial key, N, sent by the reader. Only a legitimate reader has access to the partial key N stored on the back-end database, and N will only be sent out if the tag had previously sent the correct first partial key, H(N). If the tag verifies that the reader is authentic after decrypting the forward link, the method continues at step 1110. Otherwise, the tag will not respond to any further interrogation by the apparent rogue reader.
If the tag verifies that the reader is authentic in step 1108, a secure two-way communication link is completed, and secure two-way communications can be started. This is indicated in step 1110 by the noise-encrypted communication signal, RE(X) (tag-to-reader link), and in step 1112 by the encrypted communication signal, Yˆf(R1-time pad) (reader-to-tag link) signal Y, which is encrypted by XOR'ing Y with a-function dependent on the one-time pad, R1-time pad. Backscatter communications (i.e., RE(X)) may be noise-encrypted using the encryption techniques described above in relation to FIGS. 4-10. Noise encryption in the forward link, while shown to use an XOR operation and a function of the one-time pad, R1-time pad, may alternatively use different encryption applying operations and other pseudorandom numbers besides R1-time pad. For example, the one-time pad may be modified at times (e.g., upon a request by a legitimate reader) to prevent eavesdroppers from determining, through multiple transmissions, the one-time pad and, consequently, the message contents.
Because the reader has access to both portions of the key, i.e., to H(N) and N, it has the ability to change the key values as well. Accordingly, after some elapsed time, the reader can change one or both of the values of the partial keys, H(N) and N. To perform this key value changing operation, the reader transmits both portions of the modified tag key (denoted as N′ and H(N′)) in FIG. 11, and transmits them to the tag, which stores the new values in its on-board memory. Hence, upon subsequent interrogations of the tag, the tag will have to backscatter the updated partial key, H(N′), before the reader will authenticate the tag. Assuming that the tag does, in fact, respond with the proper tag partial key, H(N′), the reader responds with the other portion of the encrypted key (N′)ˆf(R1-time pad) to establish a new secure two way communication link. This option of modifying the key values is useful in that it provides further security against a rogue reader, since a rogue reader would not see the same H(N) every time the tag is interrogated.
Referring now to FIG. 12, there is shown a timing diagram illustrating a method of establishing a secure two-way communication link between a reader and a tag of a population of tags, including applying a password lock to a singulated tag, according to an embodiment of the present invention. The password lock aspect of the invention provides security and privacy if, for example, a tag is taken out of range of a legitimate reader. In particular, using the password lock is beneficial in that once a tag is taken out of range of the reader (as happens, for example, after a customer purchases an item having a tag associated with it and leaves the store from which it is purchased), rogue readers are unable to location track the tag.
Steps 1100 through 1110 of the method in FIG. 12 relate to singulating a tag and establishing a secure two-way communication link. These steps are identical to or substantially similar to steps 1100 through 1110 in the method shown and described in relation to FIG. 11. Accordingly, the steps have been assigned the same reference numbers. Once the secure two-way communication link has been established in steps 1100 through 1110, at an appropriate time a reader issues a password lock to the singulated tag in step 1118. This password lock command, which includes a password, may be encrypted by an encryption function. In FIG. 12, this encryption is shown to be f(R1-time pad) XOR'd with the Password Lock, i.e., Password Lock ˆf(R1-time pad). Those skilled in the art will now understand that other encryption finctions may be used and that other encryption operators other than the XOR operator may be used.
To initiate communication with a tag once the tag has been password locked, the tag must first receive the correct password. Step 1120 in FIG. 12 shows the reader sending the correct password to the tag. The tag responds, at step 1122 by backscattering a noise-encrypted partial key, H(N), and one-time pad, R1-time pad, i.e., by backscattering RE(H(N), R1-time pad), identical or similar to the step 1104 describe in relation to FIG. 11 above.
Upon receipt of the backscattered signal, at step 1124 the reader consults a secure back-end database to determine whether the value of H(N) sent is valid and, accordingly, whether the tag is authentic. If the reader determines that H(N) is a valid partial key, the method continues to step 1126. Otherwise, the reader discontinues communications with the tag, assuming that it is not authentic.
If the reader verifies that the tag is authentic, at step 1126 the reader transmits the other portion of the key, N. on the forward link to the tag. According to an embodiment of the invention, N is encrypted with a function that depends on a pseudorandom number, which may be, for example, the one-time pad, R1-time pad, which was backscattered by the tag in step 1122. In FIG. 12, the encryption is shown as Nˆf(R1-time pad). Those skilled in the art will readily understand that the XOR operation is not the only operator that may be used to form the encrypted key, and that other encryption schemes may be employed.
Next, at step 1128 the tag verifies the authenticity of the reader, based on the value of the partial key, N, sent by the reader. Only a legitimate reader has access to the partial key N stored on the back-end database, and N will only be sent out if the tag had previously sent the correct first partial key, H(N), and one-time pad, R1-time pad. If the tag verifies that the reader is authentic, the method continues at step 1130. Otherwise, the tag will not respond to any further interrogation by the apparent rogue reader.
If the tag verifies that the reader is authentic in step 1128, a secure two-way communication link is completed, and secure two-way communications can be started. This is indicated in step 1130 by the noised encrypted communication signal, RE(X) (tag-to-reader link).
FIG. 13 shows how, in establishing a secure two-way communication link according to embodiments of the present invention, a rogue reader is prevented access to information backscattered by the tag. For a rogue reader to access information on the tag, it would have to initiate communication with the tag by polling and singulating the tag. This is shown as step 1140 in FIG. 13. If somehow the rogue reader succeeds at singulating the tag, at step 1142 the tag may respond by backscattering a partial key, H(N), and one-time pad, R1-time pad. The backscattered signal including the partial key, H(N), and one-time pad, R1-time pad, is shown in FIG. 13 as (H(N), R1-time pad). Upon the rogue reader receiving the backscattered signal, the only thing that it can do is send back some guess as to what the other portion of the key, N is. This is shown in step 1144 as “Nguess”. In other words, because the reader does not have access to the back-end database, it cannot determine what N is, and will have to send a guessed value of N, i.e. Nguess, optionally encrypted by some function of R1-time pad back to the tag. Because, for all practical purposes, the reader cannot guess the true value of N, the tag will not authenticate the reader and will not divulge any further information to the rogue reader. It should be mentioned that if the tag is password protected, as described above, the rogue reader will not even receive any response during polling.
FIG. 14 shows how, in establishing a secure two-way communication link according to embodiments of the present invention, a rogue tag is prevented from communicating with a legitimate reader. This security measure is important since it prevents a rogue tag from not only communicating with a legitimate reader but also from attempting to gain access to information (e.g., the other portion of key, N) stored on the back-end database through the reader. FIG. 14 shows, at step 1150, a reader initiating communication with a rogue tag by a polling signal having a random ID. Because the rogue tag has no information as to the value of a tag partial key, H(N), all that it can do is backscatter a guess, i.e., H(N)guess, at step 1152. Upon receipt of the backscattered signal, the reader consults the back-end database to verify that the tag is authentic. Because it is extremely unlikely that the rogue tag properly guessed a true value of H(N), there will be no entry in the database that corresponds to H(N). Accordingly, at step 1154 the reader will establish that the tag is a rogue tag, will not send the rogue tag the value of N, and will not communicate further with the rogue tag.
FIG. 15 shows an analog implementation of an RFID system 150, according to an embodiment of the present invention. A signal generator uses both AM and FM/PM to modulate an RF carrier signal for encrypting it. An antenna, not shown, transmits the encrypted signal. According to this embodiment, a reader 1500 includes a voltage controlled oscillator (VCO) 1501 that generates a carrier signal for broadcasting to a tag 1502. The carrier signal generated by the VCO 1501 is modulated by an analog FM/PM signal. Analog AM is also applied to the carrier by varying the gain of a variable gain amplifier (VGA) 1504. The AM and FM/PM modulated signal is transmitted to the tag 1502, which backscatter modulates the carrier signal with tag information back to the reader 1500. As described in detail above, the AM and FM/PM mask the tag information in a backscatter modulated signal. Upon receipt of the backscattered signal, a processor processes it for decryption. The modulation may be removed by applying inverse modulation. For example, the inverse of the gain applied to the transmitting VGA is applied to a receiving VGA 1506. The received signal is also mixed with the signal provided at the output of the VCO 1501 by a mixer 1503 to remove the FM/PM. Finally, the signal is sent through a demodulator 1508 to provide a baseband signal containing the tag information backscattered by the tag 1502.
FIG. 16 shows an analog implementation of an RFID system 160, in which AM is used to modulate the carrier signal, according to an embodiment of the present invention. This embodiment is similar to the embodiment shown in FIG. 15, except that no FM/PM is applied to the RF carrier signal.
FIG. 17 shows an analog implementation of an RFID system 170, in which FM/PM is used to modulate the carrier signal, according to an embodiment of the present invention. This embodiment is similar to the embodiment shown in FIG. 15, except that no AM is applied to the RF carrier signal.
FIG. 18 shows a combined analog and digital implementation of an RFID system 180, in which both AM and FM/PM are used to modulate an RF carrier signal, according to an embodiment of the present invention. This implementation is similar to the implementation shown in FIG. 15, the primary difference being that the source of signals for the AM and FM/PM are digital sources in the embodiment shown in FIG. 18. Accordingly, digital-to-analog converters (DACs) 1600 and 1602 are used to convert the digital FM/PM and digital AM signals into analog signals, respectively, before they are applied to the VCO 1501 and the gain control input of VGA 1504. A DAC 1603 is also used to convert the inverse AM to an analog signal.
FIG. 19 shows a combined analog and digital implementation of an RFID system 190, in which AM is used to modulate an RF carrier signal, according to an embodiment of the present invention. This embodiment is similar to the embodiment shown in FIG. 16, except that the source of the AM and inverse AM signals are digital. DACs 1602 and 1604 are used to convert the digital AM and digital inverse AM signal into analog signals, respectively, which control the gains of the transmitting VGA 1504 and receiving VGA 1506.
FIG. 20 shows a combined analog and digital implementation of an RFID system 200, in which FM/PM is used to modulate an RF carrier signal, according to an embodiment of the present invention. This embodiment is similar the embodiment shown in FIG. 17, except that the source of the FM/PM is digital. DAC 1600 is used to convert the digital FM/PM signal into an analog signal, which is used to modulate the VCO 1501.
FIG. 21 shows a digital implementation of an RFID system 300, according to an embodiment of the present invention. According to this embodiment, a complex noise source 1800 is converted to an analog signal by a DAC 1802. The output of the DAC 1802 is coupled to an upconverter 1804, which provides an RF carrier that is transmitted to the tag 1502. The tag 1502 backscatter modulates the carrier signal with tag information back to the reader 1500. A downconverter 1806 is configured to receive the backscatter modulated signal, which it downconverts. A complex multiplier 1810 multiplies the downconverted signal with the inverse of the complex noise signal generated by the complex noise source 1800. Alternatively, the multiplier may be an analog multiplier, in which case an inverse function 1812 is used to invert the complex noise signal, which is then applied to a DAC prior to multiplying it with the downconverted signal. Finally, a demodulator 1814 demodulates the multiplied signal to provide a baseband signal containing the tag information backscattered by the tag 1502.
While particular embodiments of the present invention have been shown and described, it will be obvious to those skilled in the art that, based upon the teachings herein, changes and modifications may be made without departing from this invention and its broader aspects. Therefore, the appended claims are intended to encompass within their scope all such changes and modifications as are within the true spirit and scope of this invention.
Patent Application
20070177738
