6 Coding and Modulation The block diagram in Figure 6.1 describes a digital communication system. Similarly, data transfer between reader and transponder in an RFID system requires three main functional blocks. From the reader to the transponder — the direction of data trans- fer — these are: signal coding (signal processing)andthemodulator (carrier circuit) in the reader (transmitter), the transmission medium (channel), and the demodulator (carrier circuit)andsignal decoding (signal processing)inthetransponder (receiver). A signal coding system takes the message to be transmitted and its signal represen- tation and matches it optimally to the characteristics of the transmission channel .This process involves providing the message with some degree of protection against inter- ference or collision and against intentional modification of certain signal characteristics (Herter and L ¨ orcher, 1987). Signal coding should not be confused with modulation, and therefore it is referred to as coding in the baseband . Modulation is the process of altering the signal parameters of a high frequency carrier, i.e. its amplitude, frequency or phase, in relation to a modulated signal, the baseband signal. The transmission medium transmits the message over a predetermined distance. The only transmission media used in RFID systems are magnetic fields (inductive coupling) and electromagnetic waves (microwaves). Demodulation is an additional modulation procedure to reclaim the signal in the baseband. As there is often an information source (input) in both the transponder and the reader, and information is thus transmitted alternately in both directions, these components contain both a modulator and a demodulator. This is therefore known as a modem (Modulator — Demodulator), a term that describes the normal configura- tion (Herter and L ¨ orcher, 1987). ReceiverTransmitter Channel Carrier circuit Carrier circuit Information source m ( t ) To information sink (user) m ( t ) Noise n ( t ) Signal processing Signal processing s ( t ) r ( t ) Figure 6.1 Signal and data flow in a digital communications system (Couch, 1997) RFID Handbook: Fundamentals and Applications in Contactless Smart Cards and Identification, Second Edition Klaus Finkenzeller Copyright  2003 John Wiley & Sons, Ltd. ISBN: 0-470-84402-7 184 6CODINGANDMODULATION The task of signal decoding is to reconstruct the original message from the baseband coded received signal and to recognise any transmission errors and flag them as such. 6.1 Coding in the Baseband Binary ones and zeros can be represented in various line codes. RFID systems normally use one of the following coding procedures: NRZ, Manchester, Unipolar RZ, DBP (differential bi-phase), Miller, differential coding on PP coding (Figure 6.2). NRZ code A binary 1 is represented by a ‘high’ signal and a binary 0 is rep- resented by a ‘low’ signal. The NRZ code is used almost exclusively with FSK or PSK modulation. Manchester code A binary 1 is represented by a negative transition in the half bit period and a binary 0 is represented by a positive transition. The Manchester code is therefore also known as split-phase coding (Couch, 1997). The Manchester code is often used for data transmission from the transponder to the reader based upon load modulation using a subcarrier. NRZ coding: Manchester coding: (bi-phase) 111 10000 111 1 000 0 1 11 10 00 0 Unipolar RZ coding: 111 1 000 0 DBP 111 1 000 0 111 1 000 0 Miller coding: Differential coding: 1 111 1 000 0 Modified Miller coding: Figure 6.2 Signal coding by frequently changing line codes in RFID systems 6.1 CODING IN THE BASEBAND 185 Unipolar RZ code A binary 1 is represented by a ‘high’ signal during the first half bit period, a binary 0 is represented by a ‘low’ signal lasting for the entire duration of the bit. DBP code A binary 0 is coded by a transition of either type in the half bit period, a binary 1 is coded by the lack of a transition. Furthermore, the level is inverted at the start of every bit period, so that the bit pulse can be more easily reconstructed in the receiver (if necessary). Miller code A binary 1 is represented by a transition of either type in the half bit period, a binary 0 is represented by the continuance of the 1 level over the next bit period. A sequence of zeros creates a transition at the start of a bit period, so that the bit pulse can be more easily reconstructed in the receiver (if necessary). Modified Miller code In this variant of the Miller code each transition is replaced by a ‘negative’ pulse. The modified Miller code is highly suitable for use in inductively coupled RFID systems for data transfer from the reader to the transponder. Due to the very short pulse durations (t pulse  T bit ) it is possible to ensure a con- tinuous power supply to the transponder from the HF field of the reader even during data transfer. Differential coding In ‘differential coding’ every binary 1 to be transmitted causes a change (toggle) in the signal level, whereas the signal level remains unchanged for a binary zero. Differential coding can be generated very simply from an NRZ signal by using an XOR gate and a D flip-flop. Figure 6.3 shows a circuit to achieve this. Pulse-pause coding In pulse-pause coding (PPC) a binary 1 is represented by a pause of duration t before the next pulse; a binary 0 is represented by a pause of duration 2t before the next pulse (Figure 6.4). This coding procedure is popular in inductively coupled RFID systems for data transfer from the reader to the transponder. Due to the very short pulse durations (t pulse  T bit ) it is possible to ensure a contin- uous power supply to the transponder from the HF field of the reader even during data transfer. Clock Data in (NRZ) Data out (differential) XOR DQ Figure 6.3 Generating differential coding from NRZ coding 186 6CODINGANDMODULATION 1 11 1 0 00 0 Pulse/Pause- length coding: START SYNC Figure 6.4 Possible signal path in pulse-pause coding Various boundary conditions should be taken into consideration when selecting a suitable signal coding system for an RFID system. The most important consideration is the signal spectrum after modulation (Couch, 1997; M ¨ ausl, 1985) and suscepti- bility to transmission errors. Furthermore, in the case of passive transponders (the transponder’s power supply is drawn from the HF field of the reader) the power sup- ply must not be interrupted by an inappropriate combination of signal coding and modulation procedures. 6.2 Digital Modulation Procedures Energy is radiated from an antenna into the surrounding area in the form of electro- magnetic waves. By carefully influencing one of three signal parameters — power, frequency, phase position — of an electromagnetic wave, messages can be coded and transmitted to any point within the area. The procedure of influencing an electromag- netic wave by messages (data) is called modulation, and an unmodulated electromag- netic wave is called a carrier. By analysing the characteristics of an electromagnetic wave at any point in the area, we can reconstruct the message by measuring the change in reception power, frequency or phase position of the wave. This procedure is known as demodulation. Classical radio technology is largely concerned with analogue modulation proce- dures. We can differentiate between amplitude modulation, frequency modulation and phase modulation, these being the three main variables of an electromagnetic wave. All other modulation procedures are derived from one of these three types. The pro- cedures used in RFID systems are the digital modulation procedures ASK (amplitude shift keying), FSK (frequency shift keying) and PSK (phase shift keying) (Figure 6.5). In every modulation procedure symmetric modulation products — so-called side- bands — are generated around the carrier. The spectrum and amplitude of the sidebands are influenced by the spectrum of the code signal in the baseband and by the modulation procedure. We differentiate between the upper and lower sideband. 6.2.1 Amplitude shift keying (ASK) In amplitude shift keying the amplitude of a carrier oscillation is switched between two states u 0 and u 1 (keying) by a binary code signal. U 1 can take on values between u 0 and 0. The ratio of u 0 to u 1 is known as the duty factor m. 6.2 DIGITAL MODULATION PROCEDURES 187 Carrier Sideband P f Figure 6.5 Each modulation of a sinusoidal signal — the carrier — generates so-called (mod- ulation) sidebands To find the duty factor m we calculate the arithmetic mean of the keyed and unkeyed amplitude of the carrier signal: ˆu m = ˆu 0 +ˆu 1 2 (6.1) The duty factor is now calculated from the ratio of amplitude change ˆu 0 −ˆu m to the mean value ˆu m : m =  ˆu m ˆu m = ˆu 0 −ˆu m ˆu m = ˆu 0 −ˆu 1 ˆu 0 +ˆu 1 (6.2) In 100% ASK the amplitude of the carrier oscillation is switched between the carrier amplitude values 2 ˆu m and 0 (On-Off keying; Figure 6.6). In amplitude modulation using an analogue signal (sinusoidal oscillation) this would also correspond with a modulation factor of m = 1 (or 100%) (M ¨ ausl, 1985). The procedure described for calculating the duty factor is thus the same as that for the calculation of the modulation factor for amplitude modulation using analogue ∆ û m û m û 1 û 0 t m = 0.5; (ASK 50%) Figure 6.6 In ASK modulation the amplitude of the carrier is switched between two states by a binary code signal 188 6CODINGANDMODULATION signals (sinusoidal oscillation). However, there is one significant difference between keying and analogue modulation. In keying, a carrier takes on the amplitude ˆu 0 in the unmodulated state, whereas in analogue modulation the carrier signal takes on the amplitude ˆu m in the unmodulated state. In the literature the duty factor is sometimes referred to as the percentage carrier reduction m  during keying: m  = 1 − ˆu 1 ˆu 0 (6.3) For the example in Figure 6.7 the duty factor would be m  = 0.66 (= 66%). In the case of duty factors <15% and duty factors >85% the differences between the two calculation methods can be disregarded. The binary code signal consists of a sequence of 1 and 0 states, with a period duration T and a bit duration τ . From a mathematical point of view, ASK modulation is achieved by multiplying this code signal u code (t) by the carrier oscillation u Cr (t). For duty factors m<1 we introduce an additional constant (1 − m), so for this case we can still multiply u HF (t) by 1 in the unkeyed state: U ASK (t) = (m · u code (t) + 1 − m) · u HF (t) (6.4) The spectrum of ASK signals is therefore found by the convolution of the code signal spectrum with the carrier frequency f Cr or by multiplication of the Fourier expansion of the code signal by the carrier oscillation. It contains the spectrum of the code signal in the upper and lower sideband, symmetric to the carrier (M ¨ ausl, 1985). A regular, pulse-shaped signal of period duration T and bit duration τ yields the spectrum of Table 6.1 (see also Figure 6.8). HF Gen 0 t Time Amplitude HF amplitude ASK modulator Digital signal HF signal T Figure 6.7 The generation of 100% ASK modulation by the keying of the sinusoidal carrier signal from a HF generator into an ASK modulator using a binary code signal 6.2 DIGITAL MODULATION PROCEDURES 189 Table 6.1 Spectral lines for a pulse-shaped modulated carrier oscillation Designation Frequency Amplitude Carrier oscillation f CR u HF · (1 − m) · (T − τ)/T 1st spectral line f CR ± 1/T u HF · m · sin(π · τ/T) 2nd spectral line f CR ± 2/T u HF · m · sin(2π · τ/T) 3rd spectral line f CR ± 3/T u HF · m · sin(3π · τ/T) nth spectral line f CR ± n/T u HF · m · sin(nπ · τ/T) 0 T t Time Amplitude Figure 6.8 Representation of the period duration T and the bit duration τ of a binary code signal 0 t Time Amplitude HF amplitude Digital signal HF signal 2FSK modulator f 2 f 1 T Figure 6.9 The generation of 2 FSK modulation by switching between two frequencies f 1 and f 2 in time with a binary code signal 6.2.2 2 FSK In 2 frequency shift keying the frequency of a carrier oscillation is switched between two frequencies f 1 and f 2 by a binary code signal (Figure 6.9). The carrier frequency f CR is defined as the arithmetic mean of the two charac- teristic frequencies f 1 and f 2 . The difference between the carrier frequency and the 190 6CODINGANDMODULATION characteristic frequencies is termed the frequency deviation f CR : f CR = f 1 + f 2 2 f CR = |f 1 + f 2 | 2 (6.5) From the point of view of the time function, the 2FSK signal can be considered as the composition of two amplitude shift keyed signals of frequencies f 1 and f 2 . The spectrum of a 2 FSK signal is therefore obtained by superimposing the spectra of the two amplitude shift keyed oscillations (Figure 6.10). The baseband coding used in RFID systems produces an asymmetric frequency shift keying: τ = T 2 (6.6) In these cases there is also an asymmetric distribution of spectra in relation to the mid-frequency f CR (M ¨ ausl, 1985). 6.2.3 2 PSK In phase shift keying the binary states ‘0’ and ‘1’ of a code signal are converted into corresponding phase states of the carrier oscillation, in relation to a reference phase. In 2 PSK the signal is switched between the phase states 0 ◦ and 180 ◦ . Mathematically speaking, the shift keying of the phase position between 0 ◦ and 180 ◦ corresponds with the multiplication of the carrier oscillation by 1 and −1. The power spectrum of a 2 PSK can be calculated as follows for a mark-space ratio τ /T of 50% (Mansukhani, 1996): P(f) =  P · T s 2  · [sin c 2 π(f − f 0 )T s + sin c 2 π(f + f 0 )T s ] (6.7) where P is transmitter power, T s is bit duration (= τ ), f 0 is centre frequency, and sin c(x) = (sin(x)/x). Sidebands P f f 2 f 1 f CR Figure 6.10 The spectrum of a 2 FSK modulation is obtained by the addition of the individual spectra of two amplitude shift keyed oscillations of frequencies f 1 and f 2 6.2 DIGITAL MODULATION PROCEDURES 191 The envelope of the two sidebands around the carrier frequency f 0 follows the function (sin(x)/x) 2 . This yields zero positions at the frequencies f 0 ± 1/T s ,f 0 ± 2/T S ,f 0 ± n/T S . In the frequency range f 0 ± 1/T S , 90% of the transmitter power is transmitted. See Figure 6.11. 6.2.4 Modulation procedures with subcarrier The use of a modulated subcarrier is widespread in radio technology. In VHF broad- casting, a stereo subcarrier with a frequency of 38 kHz is transmitted along with the baseband tone channel. The baseband contains only the monotone signal. The differ- ential ‘L–R’ signal required to obtain the ‘L’ and ‘R’ tone channels can be transmitted ‘silently’ by the modulation of the stereo subcarrier. The use of a subcarrier therefore represents a multilevel modulation. Thus, in our example, the subcarrier is first modu- lated with the differential signal, in order to finally modulate the VHF transmitter once again with the modulated subcarrier signal (Figure 6.12). In RFID systems, modulation procedures using a subcarrier are primarily used in inductively coupled systems in the frequency ranges 6.78 MHz, 13.56 MHz or 27.125 MHz and in load modulation for data transfer from the transponder to the reader. The load modulation of an inductively coupled RFID system has a similar effect to ASK modulation of HF voltage at the antenna of the reader. Instead of switching the load resistance on and off in time with a baseband coded signal, a low frequency subcarrier is first modulated by the baseband coded data signal. ASK, FSK or PSK modulation may be selected as the modulation procedure for the sub- carrier. The subcarrier frequency itself is normally obtained by the binary division of the operating frequency. For 13.56 MHz systems, the subcarrier frequencies 847 kHz (13.56 MHz ÷ 16), 424 kHz (13.56 Mhz ÷ 32) or 212 kHz (13.56 MHz ÷ 64) are usu- ally used. The modulated subcarrier signal is now used to switch the load resistor on and off. The great advantage of using a subcarrier only becomes clear when we consider the frequency spectrum generated. Load modulation with a subcarrier initially generates × 1, −1 T time Amplitude HF amplitude Digital signal HF signal 2 PSK modulator 0 f 1 t Figure 6.11 Generation of the 2 PSK modulation by the inversion of a sinusoidal carrier signal in time with a binary code signal 192 6CODINGANDMODULATION Subcarrier 212 kHz Data stream − baseband coded Carrier signal 13.56 MHz Modulated subcarrier ASK-Modulation 2 = Load modulation ASK-Modulation 1 Load modulated signal with subcarrier Figure 6.12 Step-by-step generation of a multiple modulation, by load modulation with ASK modulated subcarrier two spectral lines at a distance ± the subcarrier frequency f H around the operating frequency (Figure 6.12). The actual information is now transmitted in the sidebands of the two subcarrier lines, depending upon the modulation of the subcarrier with the baseband coded data stream. If load modulation in the baseband were used, on the other hand, the sidebands of the data stream would lie directly next to the carrier signal at the operating frequency. f Signal 0 dB −80 dB f T = 13.560 MHz f H = 212 Carrier signal of the reader, measured at the antenna coil Modulation products by load modulation with a subcarrier 13.772 MHz13.348 MHz Figure 6.13 Modulation products using load modulation with a subcarrier . communication system. Similarly, data transfer between reader and transponder in an RFID system requires three main functional blocks. From the reader to the transponder. message over a predetermined distance. The only transmission media used in RFID systems are magnetic fields (inductive coupling) and electromagnetic waves