Patent Application
20050207484
This invention relates to a method and a device for extracting the phase of a signal, possibly comprising noise, or a phase reference of a modulated signal, possibly comprising noise.
It is used particularly in applications in the domain of RFID (Radio Frequency IDentification) systems. An RFID system is usually formed from a first device containing stored data, usually called label, capable of transmitting these data in the form of radio waves at a frequency within a range usually between a few kHz and a few MHz. A second read device is then provided for acquisition of said stored data. To achieve this acquisition, the second device is capable of emitting a signal S formed from a carrier. In response to this signal S, the first device then emits another signal S′ containing said data. This other signal S′ is usually formed by a sub-carrier with a frequency proportional to the frequency of said carrier, according to a predetermined protocol. The other signal S′ may be modulated, for example using a phase modulation.
The read device should preferably be capable of detecting a phase reference of this other signal S′, so that the read device can demodulate the other signal S′. To achieve this, a number of cycles may be specifically inserted at the beginning of the other signal S′ so that the read device will have the time to detect this phase reference. One known method called the “correlation method” is currently used to do this detection type. This method consists of multiplying one or several cycles of said other signal S′, for which the phase reference is searched by one or several cycles of a reference signal S0 with a frequency identical to the frequency of said sub-carrier.
The analogue or digital results obtained from these multiplications are integrated over one or several periods. These operations are repeated regularly in time and are a maximum at the instant at which the other signal S′ and the reference signal S0 are in phase. The moment at which this maximum occurs corresponds to an instant used as the phase reference for the other signal S′.
This method has the disadvantage that it lacks precision, particularly when the other signal S′ comprises noise. It also requires the use of a large number of resources in the read device.
This invention proposes a method for estimating or measuring the phase shift of a signal S1, with a period approximately equal to a period Ts with respect to a reference signal S0 with period Ts, comprising steps consisting of receiving the signal S1 and starting a capture process including steps consisting of:
The phase shift between the reference signal S0 and the signal S1 may be estimated after step e) by a phase shift value between (j/n)*2π and ((j+1)/n)*2π, for example (j/n)*2π.
The processed signal S1 may be a digital signal, digitized on one or several bits. This signal may also be a digital signal, obtained by digitization of an analogue signal or by processing of an analogue signal.
The signal S1 may possibly be an analogue signal. In this case, a processing step of the signal S1, for example using one or several comparators or using an analogue/digital converter, may be provided before starting the capture process.
In step a), the increase in a count variable may correspond to an increment of said count variable by a fixed increment value, or by an increment value that can vary as a function of said detected state.
For example, in the case in which the signal S1 is digitized on several bits and uses a signed logic, it would be possible that the increment value is signed and that the sign of the increment value varies as a function of the state of the detected signal S1. It would also be possible for the said increment value to vary as a function of the state of the detected signal S1 and the average value or the average state of the signal S1 at the detection instant.
The increase of a count variable may also correspond to an operation other than an increment on this count variable. For example, it may consist of a multiplication of this variable by a value that depends on the state of the detected signal S1.
The capture process according to the invention may be reiterated for an integer number of periods with duration Ts. According to one variant, the capture process done in step b) and in step c) may be reiterated until one of the said count variables reaches a value Xthreshold_1. This threshold value Xthreshold_1 may be predetermined and fixed, or according to one variant it may be initialized to be equal to a predetermined value and then vary during the capture process, for example as a function of the signal-to-noise ratio relative to the signal S1 and/or the information throughput relative to the signal S1.
According to a second variant, the capture process may be reiterated until the difference in the count variable at the output from the two adjacent count means reaches a threshold Xthreshold_2.
This threshold Xthreshold_2 may be predetermined and fixed, or according to one variant it may be initialized to be equal to a predetermined value and then vary during said capture process, for example as a function of the signal-to-noise ratio relative to signal S1, or the information throughput carried by signal S1.
The invention also relates to a method of reading a transponder or a label comprising:
The signal S0 may be generated by the read device, starting from said signal with a predetermined frequency, for example by frequency division of said signal, with a predetermined frequency or using a local oscillator integrated into the read device.
The invention also relates to a device for estimating the phase shift of a signal S1 with a period approximately equal to Ts with respect to a reference signal S0 with period Ts, comprising:
The phase shift estimating device according to the invention may also comprise:
According to one particular embodiment of the device, when activated, each of the count means may be capable of detecting an instantaneous state of the signal S1 and incrementing a count variable with which it is associated among the n count variables, or leaving it unchanged, as a function of said instantaneous detected state.
The device may also include means for determining the average value or the average state M of the signal S1. Each of the count means can then be designed so that when it is activated, it detects an instantaneous state of the signal S1 and increments a count variable with which it is associated among N count variables, or leaves it unchanged, as a function of said instantaneous detected state and said average value M of the signal S1.
Means for generating the signal S0 with period TS may also be integrated into the device according to the invention. These means may include an oscillator such as a clock and/or frequency division means.
This invention will be better understood after reading the description of example embodiments given purely for information and in no way limitative, with reference to the appended figures, wherein:
Identical, similar or equivalent parts of the different figures are marked with the same numeric references so as to facilitate comparison between one figure and the next.
The different parts shown in the figures are not necessarily shown at the same scale, to make the figures more easily understandable.
An example embodiment implemented according to the invention will now be described with reference to
The signal S1 may for example be a digital signal, possibly comprising noise, with a period identical or similar to the period Ts of the reference signal S0.
The signal S1 may also for example be a digital signal with two modulated states, for example phase modulated from a carrier with a period equal to the period Ts of the reference signal S0.
The device implemented according to the invention comprises firstly control means 110 receiving the input reference signal S0, and periodically and successively activating n different count means 1000, . . . , 100n−1 within a time interval with a duration equal to the period Ts of the signal S0 and at regularly distributed instants within the said time interval.
According to one particular embodiment, means denoted 115, for example in the form of a clock such as a piezoelectric based clock, may be integrated into the said device to generate and send the reference signal S0 with period Ts, particularly at the input to control means 110.
After the signal S1 has been received at the input to said device and starting from the moment at which this device begins to carry out a phase shift measurement process of the signal S1, a given count means 100i among the n count means 1000, . . . 100n−1 may be activated, for example by control means 110 at instants
ti=t0+i*(Ts/n)+k*Ts (1).
In the expression reference (1), ‘t0’ is a moment or instant at which a process for detection of the phase of signal S1 begins. ‘i’ corresponds to the rank of the given count means 100i, which is the i(th) count means making a detection in a time interval equal to Ts, after a first count means 1000 has first made a detection within said time interval. ‘n’ is the total number of count means 1000, . . . , 100n−1, and ‘k’ is an integer corresponding to the total number of detections made by the given count means 100i.
The number n of count means may vary for example as a function of the required detection precision. According to one particular embodiment, the number n of count means may be equal to a power of 2.
Each of the different count means 1000, . . . , 100n−1 is associated with a count variable that it can modify and for which it produces the output value. The different count means 1000, . . . , 100n−1 produce the count variables x0, . . . , xn−1 respectively at the output. The operating means of each of them are identical.
When a given count means 100i (iε[0;n−1]) is activated by the control means 110 at a given time ti(iε[0;n−1]), this given count means 100i detects the state or the value of the signal S1. The given count means 100i is then capable of modifying or not modifying a given count variable xi (iε[0;n−1]) with which it is associated and for which it produces the output value, depending on the detected state of the signal S1.
The modification of the given count variable xi, if any, consists of increasing the count variable by incrementing it, for example by an increment value Q1.
Since the signal S1 is a two-state signal, according to a first variant it would be possible, for example, for a given count means 100i to increment the given count variable xi with which it is associated, by the increment value Q1 when it detects a “high” state or a logical level of the signal S1 equal to ‘1’. According to this first variant, the given count means 100i leaves the given count variable xi with which it is associated unchanged when it detects a “low” state or a logical level of the signal S1 equal to ‘0’.
According to a second variant, it would be possible for example for the given count means 100i to increment the given count variable xi with which it is associated, by the increment value Q1, when it detects a “low” state or a logical level of the signal S1 equal to ‘0’. According to this second variant, the given count means 100i leaves the given count variable xi with which it is associated unchanged when it detects a “high” state or a logical level of the signal S1 equal to ‘1’.
Since this signal S1 is a two-state signal, the count means 1000, . . . , 100n−1, may be made for example from counters formed in a given arrangement of logical gates.
Comparator means denoted 130 that can be connected to all the outputs from the different count means 1000, . . . , 100n−1 are used to compare the value of each count variable with the value of each “adjacent” or “neighbor” count variable.
A given count means 100i will be called “neighbor” or “adjacent” to another count means 100j when the said given count means is the count means that has made a detection just before or just after the other said count means 100j among all of the count means 1000, . . . , 100n−1.
We will now present an example phase detection method according to the invention performed by a device of the type mentioned above, but in this case provided with n=8 count means 1000, . . . , 1007, explicitly with reference to
The objective is to detect the phase of the signal S1 (step curve denoted C1) using the reference signal S0 (not shown but for which the period is equal to Ts).
Operation of the different count means 1000, . . . , 1007 during time is shown in
At a given start time denoted to after the signal S1 has been received at the input to said device implemented according to the invention, a process to capture the phase shift of S1 with respect to S0 begins.
Then, starting from this start moment or instant to, within a first time interval ΔT1 preferably equal to the period Ts of the reference signal S0, the said control means 110 successively activate each of the different count means 1000, . . . , 1007 from the first count means 1000 to the eighth count means 1007 once only, at moments or instants denoted to, t1, t2, t3, t4, t5, t6, t7, uniformly distributed during the said first time interval ΔT1.
The count means 1000, . . . , 1007 each perform a single detection of the signal S1 and modify or do not modify the values of their corresponding count variables, once only, within the said first time interval ΔT1.
The count means 1000, . . . , 1007, are thus used to save successive images of the state of the signal S1. Some of the count means 1000, . . . , 1007 will have detected a given state of the signal S1 during the first time interval ΔT1. Other count means 1000, . . . , 1007 will have detected another state of the signal S1.
For example, at an instant or moment denoted t3 (reference 200 on time diagram Cpt3) within the time interval ΔT1, the count means 1003 corresponding to the time diagram Cpt3 has detected a <<low >> state of the signal S1, the value of the count variable x3 with which it is associated then remains unchanged and equal to the value 0. At another instant or moment denoted t4 (reference 210 on time diagram Cpt4) within the time interval ΔT1, the count means 1004 associated with the time diagram Cpt4 has detected a <<high >> state of the signal S1, the value of the count variable x4 with which it is associated is then incremented and set equal to the value 1.
Since the first time interval ΔT1, is equal to the period Ts of the signal S0, the phase reference or the phase origin of the signal S1 is between two instants or moments tj and tj+1 (where jε[0;6]) at which the signal S1 is detected by two adjacent count means 100j and 100j+1respectively. At the end of the time interval ΔT1, the said two adjacent count means 100j and 100j+1 will have corresponding count variables xj and xj+1 with different values, one of the said two adjacent count means having detected a certain logical state of the signal for example a <<low >> state, and the other having detected another logical state of the signal, for example a <<high >> state.
Since the signal S1 may comprise noise or a may be a modulated signal possibly comprising noise, once the first time interval ΔT1 has elapsed, the operation described above is preferably reiterated.
Then, in a second time interval ΔT2 equal to Ts, the control means 110 once again activate the different count means 1000, . . . , 1007, from the first count means 1000 to the 8th count means 1007, at uniformly distributed instants in this second time interval ΔT2 and in the same order as before.
In this second time interval ΔT2, the detection instants or moments of the said two adjacent count means 100j and 100j+1 occur at times tj+Ts and tj+1+Ts respectively, and then once again enclose the phase reference of the signal S1 in time.
For example, at an instant denoted t3+Ts (reference 220 on time diagram Cpt3), the count means 1003 corresponding to time diagram Cpt3 has detected a <<low >> state of the signal S1 within the time interval ΔT2, the value of the count variable X3 with which it is associated remains unchanged and is equal to the value 0. The count means 1004 corresponding to time diagram Cpt4 has detected a <<high >> state of the signal S1 at another instant denoted t4+Ts (reference 230 on time diagram Cpt4) within the time interval ΔT2, the value of the count variable x4 with which it is associated is then incremented and set equal to the value 2.
The operation carried out during the first time interval ΔT1 and then during the second time interval ΔT2 can thus be reiterated identically over several time intervals equal to the period Ts until the time tend that will be called the end of capture time, at which the control means 110 described above stop activating the count means 1000, . . . , 1007, and at which these count means stop modifying their corresponding count variables. The total detection duration ΔT of the detection process may for example be such that:
Several variant embodiments may be provided to start the end of capture process described above.
According to a first variant, it will be possible that the capture process could for example be completed after a fixed predetermined duration after the start instant to or a time that might vary, for example as a function of several factors such as the information throughput carried by the signal S1, the signal to noise ratio for signal S1.
According to a second variant illustrated on
These means 135, for example formed by an arrangement of logical gates, can for example be used to compare the corresponding value of each of the count variables at the output from the different count means 1000, . . . , 1007, with a threshold Xthreshold_1. When a given count variable xi at the output from a given count means 100i reaches the value of the threshold Xthreshold_1, the end of capture detection means 130 emit an end of capture signal Send at the output, for example to control means 110 so that they stop activating the count means 1000, . . . , 1007, and possibly to comparator means 120 at the end of a complete cycle.
The threshold Xthreshold_1 may be a predetermined fixed value. The threshold Xthreshold_1 may also be a variable initialized to a predetermined value that is modified during the count process as a function of one or several factors, for example such as the information throughput carried by the signal S1, the signal to noise ratio related to signal S1. The throughput is defined as the number of bits per second (bits that modulate the phase, in other words that carry the information).
When the end of capture instant tend has occurred, the control means 110 stop activating the count means 1000, . . . , 1007. Comparator means 130 like those described above are then used to compare the output value of each count variable 1000, . . . , 1007 with the output value of each “adjacent” or “neighbor” count variable and to identify the pair of adjacent count means with the greatest difference in count variables.
Instants or moments tj+k1*Ts and tj+1+k1*Ts (where j and k1 are integer numbers) at which each of the count means in said pair of count means have made their detections, enclose, in time, the phase reference moment or instant. According to the previously described reference expression (1), the instants tj+k1*Ts and tj+1+k1*Ts coincide with instants t0+j*(Ts/n)+k*Ts and t0+(j+1)*(Ts/n)+k*Ts respectively.
The phase shift between the signal S0 and the signal S1 can then be estimated to be equal to a value between (j*2π)/n and ((j+1)*2π)/n), for example (j*2π)/n.
According to a third variant, the capture process described above may be completed when the difference in the count variable at the output from two adjacent count means, among all count means 1000, . . . , 1007, reaches a certain predetermined threshold Xthreshold_2.
In this third variant, the end of capture may be triggered for example by comparator means 130, or by other means provided for this purpose. The threshold Xthreshold_2 may be a fixed predetermined value. The threshold Xthreshold_2 may also be associated with a variable initialized to a predetermined value and which is modified during the count process for example as a function of one or several factors such as the information throughput carried by the signal S1, the signal to noise ratio related to signal S1.
The invention is not limited to detection of the phase of a digital signal and it can be applied to an analogue signal. Means, for example in the form of a comparator or an analogue/digital converter, may then be included in the device used according to the invention so that said analogue signal can be modified or so that processing can be done on a signal with two or more states as described above.
The invention can also be applied to phase detection of a signal comprising more than two logical states. It may be used to detect the phase of a signal S1 digitized on several bits.
The detection principle may then be similar to that described above.
When a given count means 100i (iε[0;n−1]) is activated by control means 110 at a given instant or moment ti, this given count means 100i detects a state or a logical value of the signal S1. The said given count means 100i is then capable of modifying or not modifying a given count variable xi with which it is associated and for which it produces the output value, depending on the detected state of the signal S1.
If the given count variable xi is modified, it consists of incrementing this count variable by an increment value Q2 that can vary as a function of the state of the detected signal S2. The state of the signal S2 means the state of each of the bits forming the signal S2.
This variant may be achieved for example using a device comprising a number of inputs equal to the number of bits in the signal S1. Each of the count means can then detect the state of each bit making up the signal S1.
For example, it would be possible for the count means to operate according to a signed logic:
For example, consider the case in which the signal S2 is a signal digitized on 3 bits denoted a2a1a0 where a2 is a high order bit. It would be possible according to a first variant that a given count means 100i then increments the given count variable xi with which it is associated, by a value equal to −(a1*21+a0*20), when it detects a state or a logical level of the signal S2 equal to ‘0a1a0’ and a value equal to +(a1*21+a0*20) when it detects a state or logical level of the signal S2 equal to ‘1a1a0’. The sign of the increment value Q2 may thus depend on the state of the signal S1.
In the examples described above, the count means of the device according to the invention are used such that every time that a given count means detects a signal, said count means increments a count variable with which it is associated by an increment value that may be fixed or variable depending on the state of said signal, or leaves it unchanged.
This invention is not limited to these operating examples. The count means may be designed to perform more complex operations than a simple increment or summation during each detection. This means that the difference between the different count variables can be amplified more quickly, and the duration of the detection process can be reduced.
More generally, the count means in the device according to the invention may for example be designed such that every time that a given count means detects a signal, it increases a count variable with which it is associated or it leaves it unchanged. The increase may for example be made by multiplying the variable by a number proportional to or equal to the value of the state of the said signal during the detection, or by another mathematical function that possibly depends on the state of the signal at the instant or moment of the detection.
One particular application of the invention is in the domain of RFID (Radio Frequency IDentification) systems. A phase detection device used according to the invention may for example be integrated into an RFID read device for acquisition of data within another device, for example in the form of a label or transponder, by exchange of radiofrequency signals.
An RFID read device used according to the invention can then acquire data in a label or a transponder, using the following procedure:
The read device according to the invention is capable of emitting a signal S, starting from a carrier with a given frequency.
In response to this carrier, the transponder or the label emits another phase modulated signal S1. This other signal S1 is usually formed according to predetermined protocol, from a sub-carrier with a frequency that is sub-multiple of the frequency of the said carrier emitted by the read device.
The other signal S1 may possibly comprise noise (phase skip, parasite pulses).
The said read device according to the invention is then capable of creating a signal S0 with period Ts equal to the period of said sub-carrier from which the signal S1 emitted by the label of the transponder is formed, so that the read device can read variations of the other signal S1 emitted by the label or the transponder. For example, the signal S0 may be formed from one or several piezoelectric based clock as was mentioned above or by frequency division of the carrier, using frequency division means. Thus, the frequency of the sub-carrier starting from which the other signal S1 is formed, is known to the read device. The phase origin or the phase reference of the other signal S1 remains to be determined, so that the other signal can be demodulated.
In order to determine this phase origin, an estimate of the phase shift between the reference signal S0 and the signal S1 can then be made using a process like that described above. This phase shift estimate can then be used to generate a signal S′0 with a period equal to the period Ts of the sub-carrier and synchronized with S1 so as to determine the phase modulations of S1 that carries the information.
This process could begin as soon as the read device receives the signal S1.
| Number | Date | Country | Kind |
|---|---|---|---|
| 04 50534 | Mar 2004 | FR | national |
