power reduction in rfid

power reduction in rfid - NTT DoCoMo Technical Journal Vol....

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: NTT DoCoMo Technical Journal Vol. 8 No.1 Technologies to Reduce Power Consumption of Active RFID Readers Shinzo Ohkubo and Kosei Takiishi Two technologies and their effectiveness are described, devised with the aim of reducing the power consumption required for mounting an active RFID reader on a mobile terminal. 1. Introduction New types of services are expected in the mobile ubiquitous environment [1], where all sorts of objects and computers exchange information either directly or through networks. In such an environment, Radio Frequency IDentification (RFID) is attracting interests in recent years as one of the realizable technologies for providing connectivity that includes various objects. RFID systems consist of a radio tag for transmitting an ID and a reader for reading the ID, and have been utilized as a technology mainly for performing product identification and management by attaching a tag on an object. Classified broadly, there are two types of RFID systems: passive types, in which a tag transmits an ID using electrical current induced in its antenna by the incoming radio frequency signal transmitted from a reader; and active-types, where a tag contains its own battery and can transmit an outgoing ID signal without the need for an external power source. Active types have the feature of longer communication distances compared with passive types, even with a hand-held reader having restriction on battery capacity and antenna diameter, because the active type includes a battery mounted on the tag. Due to this difference in communication distances, serviceability may be characterized as follows. With the passive type having a shorter communication dis- 33 tance, a user can obtain an ID by intentionally holding a reader to power consumption, even after chip integration is performed. against a tag. That is, in order to initiate data transmission, user Accordingly, new technology with the aim of reducing power intention is required. On the other hand, since the active type consumption is required. has a communication range of several meters, a user carrying a This article describes the devised technologies for reducing reader can obtain an ID simply by approaching a tag. That is, to the power consumption of a reader, with the goal of mounting receive data from an active RFID tag, the intention of a user is an active RFID reader on a mobile terminal. not necessarily required. Accordingly, with the active type, it is possible to realize data transmission service either by supporting a user-initiated action or after a user is given an “awareness” by a reader that has received an ID from a nearby tag (Figure 1). 2. Technologies for Reducing Power Consumption of a Reader *1 Here we describe the TimeHopping (TH) access/intermittent reception control method, and multiple zero-crossing In a mode with the reader placed in a fixed location and a *2 demodulation using an under-sampling technique below. user carrying an object with a tag, however, the restriction on the communication distance can be eased by the user and tag 2.1 TH Access/Intermittent Reception Control Method approaching the reader, a network for providing feedback to the Reduced power consumption of the monitoring system ser- user of the ID obtained by the reader in the fixed location is vice can be realized by intermittent reception control, which required, resulting in a service expansion that is considered to performs reception operations only at the time when an ID is be somewhat limited. This means that a mode in which a user transmitted from a tag. To perform intermittent reception opera- carries the reader is essential, and if the reader function can be tions, one method is for tags to periodically transmit the ID. mounted on a mobile terminal, user-friendliness is further However, in this case, the reader tends to incur continual recep- improved. tion failures due to collisions caused by the series of signals Figure 2 shows examples of service using an active RFID reader mounted on a mobile terminal. The monitoring system received from a number of various tags. Accordingly, it is important to reduce the number of continual reception failures. service shown in Fig. 2 (a) requires a reader to continue moni- This technology uses an access/intermittent reception con- toring an ID, while the detection system service shown in Fig. 2 trol method in which a TH sequence is generated based on an (b) requires a reader to detect an ID. ID, and individual tag transmission timing is randomized Currently, commercially-available active RFID reader according to the generated TH sequence that causes a reduction deemed to be difficult to be mounted on a mobile terminal due in the continual collision rate, while a reader performs intermit- Mobile network Tag Key Stores and points of interest* Places Object Tag Wallet Tag Bag User Mobile terminal with active RFID reader Trains* *Tags are placed inside the store, store entrances, and inside the train. About 10 meters maximum (as a radius from the user) Figure 1 Mobile terminal with active RFID reader *1 TH: A method of changing transmission timings of a signal based on a pattern determined by a code sequence. *2 Under-sampling: A sampling method which is used for converting signals of a high frequency range into a low frequency range, and performs sampling operations using a frequency lower than the carrier frequency, and equal to or greater than twice the one side of the frequency bandwidth transmission signal. 34 NTT DoCoMo Technical Journal Vol. 8 No.1 “Bag” is left behind. “This tower is . . . .” ¥ Tag 【Man navigating and reporting positional information】 (supplementing GPS within a building, underground shopping areas, etc.) 【Valuables management】 (Reporting items left behind, searching for items) 【Providing spot information】 Closed to traffic! (a) Monitoring system service (b) Detection system service Tag 【Communication switching according to a user’s location】 【Dangerous area report】 Tag With vibration sensor Incoming call control 【Reports that an object has moved】 Switched to manner mode 【Mobile function control according to a user’s location】 Figure 2 Service examples tent reception operations. The principle for generating the TH sequence is described Elements of GF (23) Power Polynomial represenequation tation representation k1, k2 below. As an example, generation of a TH sequence with a cycle of 7 is shown in F igure 3 . Two blocks ( k 1, k 2) are 000 0 0     0 obtained by selecting three sequential bits from the Least 001 1 1     1 010 2 α   α 011 3 2 α α 100 4 α3   α+ 1 in Formula a. Both k1 and k2 are prime power representations 101 5 α4 α2 + α corresponding to each element of k1 and k2. 110 6 α5 α2 + α+ 1 111 7 α6 α2  + 1 MSB LSB Tag ID (128 bits) *3 Significant Bit (LSB) side of an ID. Each block is converted to *4 3 the elements of a Galois field GF (2 ) to obtain the coefficients ^ 2 ^ k2 ^ ^ k1 a P(x) = x +α x +α By sequentially substituting x in Formula a with an ele2 6 3 ment {1,α,α, ···,α} of GF (2 ), a TH sequence {Sj} is derived with values obtained by calculations that follow a rule such that ・・・ 1 1 0 0 1 1 ・・ k2 k1 2 k1 = 3 α2 k2 = 6 α5 P(x)= x 2 +α5x +α2 α3 =α+ 1 x = 1) P(1) = 1 +α5 +α2 =1 + (α2 +α + 1) +α2 =α 20 x =α) P(α) =α2+α6 +α2 =α2 + (α2 + 1) +α2 = α2 + 1 71 : :: x =α6) P(α ) =α12 +α11 +α2 =·························= α2 + 1 76 TH sequence: Sj}j { 6 Figure 3 Example of generating TH sequence (cycle 7) *3 LSB: Least significant bit. Most significant bit is MSB. *4 Galois field: A field that contains a finite number of elements, and which is closed under the four basic arithmetic operations, such as {0, 1, 2, 3, 4}; it can be considered as the set of remainders obtained by dividing integers by a prime number (e.g., 5). It is commonly used in coding theories, etc. 35 the sum of items having the same power index equals zero (for 2 2 example,α +α = 0, 1 + 1 = 0). of receptions. In order to receive a detection system service simultaneous- Figure 4 shows an example of TH transmission by a tag ly, the reader performs continuous reception operations to and the reception operation of a reader. Assigning a value Sj to receive IDs for detection system services. The tag transmission the generated TH sequence as the slot number to be transmitted, frequency for the detection system service is different from that the tag transmits an ID and the phase information j of the TH of the monitoring system service. sequence. The reader performs continuous reception operations until the first ID is received by monitoring. The TH sequence is 2.2 Multiple Zero-Crossing Demodulation using Under-Sampling generated based on the received ID in the same way as in the tag. Furthermore, subsequent ID reception timing is estimated This technology is multiple zero-crossing demodulation, based on the phase information [2]. In this way, intermittent which deals with degradation of reception quality level caused reception operations are performed only at the estimated timing by frequency shifts and enables both the simplification of reception equipment structure as well as digitization to reduce power Tag (Detection system) consumption. A zero-crossing demodulator detects a binary *5 symbol by first detecting the direction of a Frequency Shift Example of TH sequence (2, 7, 4, 4, 2, 3, 7) Transmission of ID + Phase information j *6 Keying (FSK) signal prior to being converted into a carrier *7 frequency band crossing In-phase/Quadrature-phase (I/Q) axis, Tag (Monitoring 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 system) Slot # ON Reader OFF of the level change of each of the I and Q signals (Figure 5). t Intermittent Start that is, the direction (from High-to-Low or from Low-to-High) Intermittent However, as the frequency shift becomes larger, the phase tran- Continuous sition of the FSK signal within one symbol period equal to or Continuous more than π [rad] does not exist. As a result, crossover of the /2 I/Q axis disappears (zero-crossing disappearance), causing the Figure 4 Example of TH transmission by tag, and reception operation of reader increase of symbol errors. Therefore, multiple zero-crossing Q0 In BPF IN OUT fIF Q0 Limiter IN OUT I0 ⑥③ Zero-crossing demodulator Delay 1 IN OUT Delay 2 Q1 I1 ⑤ ② Zero-crossing demodulator Counter (LPF) Code deter- Out mination Phase transition of reduced FSK signal Q1 Delay 1 Delay 1: 2 = 1/4fIF τ Delay 2:τ= 1/8fIF H : High level L : Low level : From Low to High : From High to Low I0 ⑦ ④ Σ IN OUT fs ⑧ ① I1 Crossing direction I Q Symbol ① ② ③ ④ H 0 L 1 H 1 ⑤ H ⑥ L ⑦ H ⑧ L 1 0 0 1 L 0 Figure 5 Configuration of receiver (demodulator and perimeter) *5 Symbol: In this article, the smallest unit of data to be transmitted. One symbol consists of n bits (where n is a positive integer). *6 Frequency Shift Keying: A digital modulation method to transmit a digital signal by associating it with different frequencies (speeds of different phase transitions). *7 I/Q: In-phase and quadrature component of the complex digital signal. 36 NTT DoCoMo Technical Journal Vol. 8 No.1 demodulators are provided to reduce the number of zero-crossing disappearances by providing multiple I/Q axes of different angles. The frequency shift is caused by the fact that the fre- 3. Prototype External RFID for Mobile Terminal quency accuracy of an oscillation system using a Surface *8 An exterior view of prototype active RFID reader is shown Acoustic Wave (SAW) resonator is low. However, since SAW in Photo 1, and basic specifications are shown in Table 1. To resonators are widely used for commercial tags and weak-out- not only evaluate the effectiveness of the devised technologies put radio modules, their use is necessary to enable the miniatur- for reducing power consumption, but also to enable actual expe- ization and cost reduction of tags. rience of the services as shown in Fig. 2, a prototype RFID A possible configuration of a receiver applying this technol- reader was equipped with an interface and shape capable of ogy using the case of a multiplication factor of 2 is shown in external attachment to a mobile terminal (N902i). In addition, a Fig. 5. A received signal with the frequency converted into an dielectric chip antenna *9 Intermediate Frequency (IF) range is input to a limiter through *14 was set on the circuit board of the reader. *10 a Band Pass Filter (BPF) . The carrier frequency of an active RFID reader is generally in the 300 MHz range, and since it is difficult to produce a small-sized BPF having the required bandwidth (Formula s) in this frequency range, frequency conversion to an IF range is required. Multiple zero-crossing demodulation occurs by performing under-sampling of the FSK signal at different and multiple timings after conversion into a binary signal by the limiter [3]. With this configuration, since binary conversion is performed within the IF range, unlike conventional configurations [4] [5], a digital configuration which requires neither Automatic Gain Control (AGC) *11 nor an analog-to-digi- tal converter can be used after the IF range, enabling reduced Photo 1 Exterior view of prototype RFID reader power consumption. In order to prevent aliasing *12 of unnecessary frequencies Table 1 Basic specifications of prototype RFID reader from being generated (other than the received signal by under- Frequency 314 MHz, 315 MHz, 2 channels sampling), sampling frequency fs must satisfy Formula s. Transmission power Weak (Radio Law 4-1) Transmission speed *13 is the divider ratio integer of 1 or greater, and BW is the band- width of the BPF. In addition, time difference τ of samplings is given in Formula d, where M is the multiplication factor. Binary FSK/Asynchronous demodulation (section 2.2) Encoding type NRZ ID 128 bits Transmission packet structure ID + Phase Information (3 bits) + CRC (16 bits) Applied with FEC Packet Length: 26.7 ms (at 9.9 kbit/s) Size, weight In this formula, fIF is the central frequency of the IF range, N 10 meters, maximum (at 9.9 kbit/s) Modulation/ demodulation s fs = fIF /N>BW 9.9 kbit/s, 99 kbit/s to be selected Access distance 45 × 89 × 16 mm, 45 g d Cyclic Redundancy Check (CRC): Error detection method capable of detecting errors generated during data transmission. Forward Error Correction (FEC): A system where redundant information is added on the transmitting side, and errors generated during data transmission are corrected using the redundant information on the receiving side. *8 SAW resonator: An oscillator circuit method for a radio module integrated as a high-frequency component of radio equipment. Mainly, there are the crystal method and SAW resonator method for oscillator circuits, and a SAW resonator is used because of low cost, even though frequency stability is somewhat inferior to that of the crystal method. *9 IF: A frequency that a signal of multiple frequency channels is converted into during demodulation. By making this frequency lower than the carrier frequency range, a filter for the required bandwidth can be easily constructed, so that only the required signal is selected and extracted. *10 BPF: A filter allowing only a specific frequency band to pass. τ = 1/4 MfIF 37 nologies, battery life is about 30 hours for 1 tag, about 21 hours 4. Effectiveness for 5 tags, about 16 hours for 10 tags, and about 12.5 hours for The effectiveness of the two devised technologies evaluated 20 tags [6]. The trial unit incorporates Field Programmable Gate *15 Arrays (FPGAs) , therefore about 25% of power consumption by the prototype RFID is described below. during packet reception is consumed during wait-time states, 4.1 TH Access/Intermittent Reception Control Method with the above results. However, since power consumption dur- Figure 6 shows the voltage variation characteristics of the ing wait-time can be reduced after chip integration, it is antici- reader battery and Table 2 shows the continuous collision gen- pated that further power reduction effects can be expected. eration rates for an ID. The tag settings were average transmis- 2) Effectiveness of Reducing Continuous Collision rates sion interval of 1 second, TH sequence cycle of 7, and transmis- Table 2 shows the average value of continuous collision sion speed of 9.9 kbit/s for each ID. Since collision frequency rates obtained for each tag. For this purpose, a continuous colli- was being evaluated, the transmission speed was set to 9.9 sion is defined as a condition when measurements are attempted kbit/s to provide a longer transmission packet length. A speed of 100 times in 30 seconds from the time an intermittent reception 9.9 kbit/s is applicable for services requiring an access distance operation is started in the reader after the continuous reception of about 10 meters. of an ID by the monitoring system service (Fig. 4), and no ID is 1) Effectiveness of Power Consumption Reduction received during the attempt. Since differences between trans- As seen from Fig. 6, it is clear that battery life can be extend- mission start timings of different tags are important, the timings ed depending on the number of tags with the devised technolo- were changed for each trial. Furthermore, reception power from gies applied as compared to continuous reception. While a bat- each tag was great enough to ignore a no-reception condition tery lasts for about 9 hours regardless of the number of tags in caused by ambient noise. It is clear that continuous collision the case of continuous reception, by applying the devised tech- generation rates can be reduced when the devised technologies are applied. In the case of periodic transmissions, the probabilities of no ID reception for a 30-second duration are about 10% Voltage (V) 4 for 5 tags, about 20% for 10 tags, and about 40% for 20 tags. With respect to the above, the devised technologies over 100 tri- 3 als reduced this probability to 0% for 5 tags, about 0.2% for 10 Number of tags 20 = 1, 5, 10, 20 2 10 5 1 tags, and about 2% for 20 tags. Intermittent reception (Devised technology) 1 4.2 Multiple Zero-Crossing Demodulation using Continuous reception 0 0 5 Under-Sampling Technique 10 15 20 Measurement time (hours) 25 30 Quantitative valuations of power consumption reductions due to the simplification of the configuration and digitization Figure 6 Voltage variation characteristics of battery in trial unit of RFID reader will be clarified further after actual chip integration. In this section, the effectiveness regarding improvements of resistance Table 2 Continuous collision generation rates of Tag ID Number of tags (Units) 5 10 20 TH transmission (Devised technology) No occurrence during measurements 0.2 1.65 Periodic transmission (Conventional) 9.8 18.9 40.2 *11 Automatic Gain Control: A function to automatically control the amplitude of an output signal to remain constant. *12 Aliasing: A distortion caused by a part of a sampled signal being mixed with a frequency component of the original signal when performing sampling with a frequency lower than twice the one side of the frequency bandwidth of a transmission signal. 38 properties against possible frequency shifts will be described. Required reception power characteristics to prevent frequency shifts are shown in Figure 7. Communication speed was set to 99 kbit/s, which is more vulnerable to frequency shifting than a rate of 9.9 kbit/s. The required reception power is defined as *13 Divider ratio: Ratio obtained when a clock frequency or the like is converted into a frequency obtained by dividing the original frequency by an integer. *14 Chip antenna: A type of antenna with a flat and rectangular shape. Widely used with a mobile terminal due to its miniaturization and light weight. *15 FPGA: A large-scale integrated circuit capable of being rewritten, consisting of cells arranged in the shape of an array and wiring elements. 1 –90 M=1 2 3 5 ID no-reception rate Required reception power value (dBm) NTT DoCoMo Technical Journal Vol. 8 No.1 –100 –110 0 10 20 30 Frequency shift (kHz) 40 50 10–1 ID_A ID_B 10–2 10–3 –113 –110 Reception power (dBm) –105 Figure 8 ID no-reception rate characteristics Figure 7 Required reception power characteristics against frequency shift considered as the reason for generating a difference in the level reception power measured at the output end of the reception –2 of reception quality between ID_A and ID_B. However, it is antenna when the ID no-reception rate becomes 10 . It is clear believed that this difference can be further lessened by adjust- that resistance properties against frequency shifts are improved ment of the LPF after demodulation. As described above, even as the multiplication factor M is increased. Even in the case of a when Manchester encoding multiplication factor of 2, with respect to –108.4 dBm of reception quality levels can be obtained even with Non-Return required reception power at the time of no frequency shifts, the to Zero (NRZ) required reception power with frequency shifts of +/–100 ppm cy shift keying of the tag’s SAW generator to not be unneces- (parts per million) (+/–31.5 kHz at 315 MHz) is –106.2 dBm, sarily enlarged. As a result, it is possible to assign the entire with the resulting required reception power being suppressed by realizable frequency shift keying to transmission. When the about 2 dBm. amount of phase transition with Manchester encoding is the Furthermore, since sampling is performed in the IF range using this demodulation method, the pattern effect *16 *18 *19 and the like are not used, uniform encoding. NRZ encoding enables the frequen- same as NRZ encoding, communication speeds for NRZ encod- can be ing can be twice those of Manchester encoding. Accordingly, avoided, resulting in realization of a uniform reception quality the time required for transmitting an ID is reduced by one-half, level uninfluenced by ID bit patterns. resulting in reduced collision rates and lengthened battery life. Figure 8 shows ID no-reception rate characteristics for an ID consisting of symbol 0 and 1 being repeated alternately 5. Conclusion (hereinafter referred to as ID_A), and for an ID consisting of the In this article, technologies devised aiming at reduction of same symbol for all 128 bits (ID_B). Communication speed power consumption required for mounting an active RFID read- used during trials was 99 kbit/s, as before. The frequency com- er on a mobile terminal have been described. Based on evalua- ponents of ID_A are concentrated around the frequencies at tion results of a prototype RFID reader, it is expected that which 1 and 0 are alternately repeated, while the frequency power consumption can be reduced without degrading the level components of ID_B are concentrated around 0 Hz due to no of reception quality. With the monitoring system service, power voltage change as a result of the same symbol being repeated. consumption can be further reduced by application of intermit- Accordingly, the high frequency components of ID_A are tent reception. After chip integration, it is expected that a reader greater than those of ID_B, resulting in the Low Pass Filter applying this technologies can be mounted on a mobile terminal (LPF) *17 passing less power for ID_A than for ID_B. This is *16 Pattern effect: Level fluctuation occurring when the same symbols are continued in a transmission system after the direct current is powered off. If level fluctuation is generated, bit errors are increased. *17 LPF: A filter allowing a low frequency band to pass. without reducing battery life. *18 Manchester encoding: A method of transmitting symbols in which “0” is transmitted as a pulse level change from high to low, and “1” is transmitted as a level change from low to high. Twice the bandwidth is required as compared with NRZ encoding. *19 NRZ: A method of transmitting symbols such that a pulse is transmitted for each symbol, using voltage values corresponding to “0” as low and “1” as high. 39 References Shinzo Ohkubo [1] K. Imai et al.: “A New Direction in 4G Infrastructure Research—Growth Senior Research Engineer, Wireless Laboratories into a Ubiquitous World—,” NTT DoCoMo Technical Journal, Vol. 6, No. 3, pp.4–15, Dec. 2004. [2] K. Takiishi, S. Ohkubo and H. Suda: “Random Access and Intermittent Joined in 1993. Engaged in research and development on FLEX-TD systems, and Fourth-Generation mobile communication systems. A member of IEICE. Reception with Time Hopping,” Proc. of IEICE Society Conference, B-5204, Sep. 2003 (In Japanese). [3] S. Ohkubo and H. Suda: “Multi-Zero-Crossing Demodulator Using an Under-sampling Technique,” Proc. of IEICE General Conference, B-5- Kosei Takiishi 205, Mar. 2004 (In Japanese). Wireless Laboratories [4] E. K. B. Lee and H. M. Kwon: “NEW BASEBAND ZERO-CROSSING DEMODULATOR FOR WIRELESS COMMUNICATIONS, PART-I: PERFORMANCE UNDER STATIC CHANNEL, ” Proc. of IEEEMILCOM, pp. 543–547, 1995. [5] S. Samadian, R. Hayashi and A. A. Abidi: “Demodulators for a Zero-IF Bluetooth Receiver,” IEEE J. Solid-State Circuits, Vol. 38, No. 8, pp. 1393–1396, 2003. [6] K. Takiishi, S. Ohkubo, T. Sugiyama and N. Umeda: “ Power Consumption Reduction by Random Access and Intermittent Reception with Time Hopping,” Proc. of IEICE Society Conference, B-5-101, Sep. 2005 (In Japanese). 40 Joined in 2001. Engaged in research and development on Fourth-Generation mobile communication systems. A member of IEICE. ...
View Full Document

Ask a homework question - tutors are online