Modeling and Simulation of New Encoding Schemes for High-Speed UHF RFID Communication

Similar documents
Rejection of PSK Interference in DS-SS/PSK System Using Adaptive Transversal Filter with Conditional Response Recalculation

Performance Analysis of Multi User MIMO System with Block-Diagonalization Precoding Scheme

A study of turbo codes for multilevel modulations in Gaussian and mobile channels

The Performance Improvement of BASK System for Giga-Bit MODEM Using the Fuzzy System

Revision of Lecture Twenty-One

Parameter Free Iterative Decoding Metrics for Non-Coherent Orthogonal Modulation

Digital Transmission

Space Time Equalization-space time codes System Model for STCM

High Speed, Low Power And Area Efficient Carry-Select Adder

Calculation of the received voltage due to the radiation from multiple co-frequency sources

A High-Sensitivity Oversampling Digital Signal Detection Technique for CMOS Image Sensors Using Non-destructive Intermediate High-Speed Readout Mode

Multipath Propagation. Outline. What is OFDM? (OFDM) for Broadband Communications and. Orthogonal Frequency Division Multiplexing

Efficient Large Integers Arithmetic by Adopting Squaring and Complement Recoding Techniques

antenna antenna (4.139)

High Speed ADC Sampling Transients

Comparative Analysis of Reuse 1 and 3 in Cellular Network Based On SIR Distribution and Rate

Walsh Function Based Synthesis Method of PWM Pattern for Full-Bridge Inverter

Resource Allocation Optimization for Device-to- Device Communication Underlaying Cellular Networks

Phasor Representation of Sinusoidal Signals

FFT Spectrum Analyzer

Adaptive Modulation for Multiple Antenna Channels

Uncertainty in measurements of power and energy on power networks

Figure.1. Basic model of an impedance source converter JCHPS Special Issue 12: August Page 13

AN IMPROVED BIT LOADING TECHNIQUE FOR ENHANCED ENERGY EFFICIENCY IN NEXT GENERATION VOICE/VIDEO APPLICATIONS

MIMO-OFDM Systems. Team Telecommunication and Computer Networks, FSSM, University Cadi Ayyad, P.O. Box 2390, Marrakech, Morocco.

Topology Control for C-RAN Architecture Based on Complex Network

THE USE OF CONVOLUTIONAL CODE FOR NARROWBAND INTERFERENCE SUPPRESSION IN OFDM-DVBT SYSTEM

A Spreading Sequence Allocation Procedure for MC-CDMA Transmission Systems

PRACTICAL, COMPUTATION EFFICIENT HIGH-ORDER NEURAL NETWORK FOR ROTATION AND SHIFT INVARIANT PATTERN RECOGNITION. Evgeny Artyomov and Orly Yadid-Pecht

Research of Dispatching Method in Elevator Group Control System Based on Fuzzy Neural Network. Yufeng Dai a, Yun Du b

熊本大学学術リポジトリ. Kumamoto University Repositor

Hardware Design of Filter Bank-Based Narrowband/Wideband Interference Canceler for Overlaid TDMA/CDMA Systems

Bit Error Probability of Cooperative Diversity for M-ary QAM OFDM-based system with Best Relay Selection

NATIONAL RADIO ASTRONOMY OBSERVATORY Green Bank, West Virginia SPECTRAL PROCESSOR MEMO NO. 25. MEMORANDUM February 13, 1985

Chaotic Filter Bank for Computer Cryptography

Index Terms Adaptive modulation, Adaptive FEC, Packet Error Rate, Performance.

Throughput Maximization by Adaptive Threshold Adjustment for AMC Systems

Research on Peak-detection Algorithm for High-precision Demodulation System of Fiber Bragg Grating

To: Professor Avitabile Date: February 4, 2003 From: Mechanical Student Subject: Experiment #1 Numerical Methods Using Excel

PAPER Effect of Joint Detection on System Throughput in Distributed Antenna Network

The Application of Interpolation Algorithms in OFDM Channel Estimation

On Timing Offset and Frequency Offset Estimation in LTE Uplink *

Side-Match Vector Quantizers Using Neural Network Based Variance Predictor for Image Coding

NOVEL ITERATIVE TECHNIQUES FOR RADAR TARGET DISCRIMINATION

Performance Analysis of Power Line Communication Using DS-CDMA Technique with Adaptive Laguerre Filters

Bit-interleaved Rectangular Parity-Check Coded Modulation with Iterative Demodulation In a Two-Node Distributed Array

Research on Controller of Micro-hydro Power System Nan XIE 1,a, Dezhi QI 2,b,Weimin CHEN 2,c, Wei WANG 2,d

DVB-T/H Digital Television Transmission and its Simulation over Ricean and Rayleigh Fading Channels

Impact of Interference Model on Capacity in CDMA Cellular Networks. Robert Akl, D.Sc. Asad Parvez University of North Texas

A Novel Soft-Switching Two-Switch Flyback Converter with a Wide Operating Range and Regenerative Clamping

Uplink User Selection Scheme for Multiuser MIMO Systems in a Multicell Environment

Near-Optimum Blind Decision Feedback Equalization for ATSC Digital Television Receivers

A High-Speed Multiplication Algorithm Using Modified Partial Product Reduction Tree

The Spectrum Sharing in Cognitive Radio Networks Based on Competitive Price Game

On the Feasibility of Receive Collaboration in Wireless Sensor Networks

Approximating User Distributions in WCDMA Networks Using 2-D Gaussian

MTBF PREDICTION REPORT

A NOVEL PREAMBLE DESIGN FOR CHANNEL ESTIMATION IN MIMO- OFDM SYSTEMS RESULTING IN ENHANCED THROUGHPUT

Multicarrier Modulation

Available online at ScienceDirect. Procedia Computer Science 93 (2016 )

A new family of linear dispersion code for fast sphere decoding. Creative Commons: Attribution 3.0 Hong Kong License

A Preliminary Study on Targets Association Algorithm of Radar and AIS Using BP Neural Network

Exponential Effective SIR Metric for LTE Downlink

California, 4 University of California, Berkeley

A Current Differential Line Protection Using a Synchronous Reference Frame Approach

Error Probability of RS Code Over Wireless Channel

A Digital Content Distribution Using a Group-Key and Multi-layered Structure Based on Web

Prevention of Sequential Message Loss in CAN Systems

IEE Electronics Letters, vol 34, no 17, August 1998, pp ESTIMATING STARTING POINT OF CONDUCTION OF CMOS GATES

Novel Sampling Clock Offset Estimation for DVB-T OFDM

MINIMUM OVERHEAD BURST SYNCHRONIZATION FOR OFDM BASED BROADBAND TRANSMISSION. Michael Speth, Dirk Daecke, Heinrich Meyr

Keywords LTE, Uplink, Power Control, Fractional Power Control.

TECHNICAL NOTE TERMINATION FOR POINT- TO-POINT SYSTEMS TN TERMINATON FOR POINT-TO-POINT SYSTEMS. Zo = L C. ω - angular frequency = 2πf

ETSI TS V8.4.0 ( )

A NSGA-II algorithm to solve a bi-objective optimization of the redundancy allocation problem for series-parallel systems

Performance of WCDMA Downlink FDD Mode at 10 MHz Bandwidth

Effect of Time-Interleaved Analog-to-Digital Converter Mismatches on OFDM Performance

RC Filters TEP Related Topics Principle Equipment

DESIGN OF OPTIMIZED FIXED-POINT WCDMA RECEIVER

The Effect Of Phase-Shifting Transformer On Total Consumers Payments

CELL SEARCH ROBUST TO INITIAL FREQUENCY OFFSET IN WCDMA SYSTEMS

Network Reconfiguration in Distribution Systems Using a Modified TS Algorithm

Markov Chain Monte Carlo Detection for Underwater Acoustic Channels

A New Type of Weighted DV-Hop Algorithm Based on Correction Factor in WSNs

Evaluate the Effective of Annular Aperture on the OTF for Fractal Optical Modulator

A High Noise Immunity, Channel Finger Touch Sensing IC Using OFDM and Frequency Translation Technique

STUDY ON LINK-LEVEL SIMULATION IN MULTI- CELL LTE DOWNLINK SYSTEM

Performance Study of OFDMA vs. OFDM/SDMA

Energy Efficiency Analysis of a Multichannel Wireless Access Protocol

Inverse Halftoning Method Using Pattern Substitution Based Data Hiding Scheme

Encoding of passive anticollision radio-frequency identification surface acoustic waves tags

1 GSW Multipath Channel Models

Mismatch-tolerant Capacitor Array Structure for Junction-splitting SAR Analog-to-digital Conversion

LETTER An MMSE-Nulling Partial-PIC Receiver for Multiuser Downlink MIMO MC-CDMA Systems

Low Switching Frequency Active Harmonic Elimination in Multilevel Converters with Unequal DC Voltages

Ergodic Capacity of Block-Fading Gaussian Broadcast and Multi-access Channels for Single-User-Selection and Constant-Power

Signal reconstruction in ANS-DM decoder

Energy Efficient Adaptive Modulation in Wireless Cognitive Radio Ad Hoc Networks

MIMO Precoding Using Rotating Codebooks

Passive Filters. References: Barbow (pp ), Hayes & Horowitz (pp 32-60), Rizzoni (Chap. 6)

Transcription:

Modelng and Smulaton of New Encodng Schemes for Hgh-Speed UHF RFID Communcaton Sang-Hyun Mo, J-Hoon Bae, Chan-Won Park, Hyo-Chan Bang, and Hyung Chul Park In ths paper, we present novel hgh-speed transmsson schemes for hgh-speed ultra-hgh frequency (UHF) radofrequency dentfcaton communcaton. For hgh-speed communcaton, tags communcate wth a reader usng a hgh-speed Mller (HS-Mller) encodng and multple antennas, and a reader communcates wth tags usng extended pulse-nterval encodng (E-PIE). E-PIE can provde up to a two-fold faster data rate than conventonal pulse-nterval encodng. Usng HS-Mller encodng and orthogonal multplexng technques, tags can acheve a two- to three-fold faster data rate than Mller encodng wthout degradng the demodulaton performance at a reader. o verfy the proposed transmsson scheme, the MALAB/Smulnk model for hgh-speed backscatter based on an HS-Mller modulated subcarrer has been desgned and smulated. he smulaton results show that the proposed transmsson scheme can acheve more than a 3 db hgher BER performance n comparson to a Mller modulated subcarrer. Keywords: RFID, passve tag, hgh-speed Mller encodng, mult-antenna, extended pulse-nterval encodng. Manuscrpt receved Aug. 9, 24; revsed Jan. 29, 25; accepted Feb. 3, 25. hs work was supported by ERI R&D Program [5ZC2, he Development of passve RFID technology for mprovement of relablty and hgh speed wth large scaled memory] funded by the Government of Korea. Sang-Hyun Mo (shmo@etr.re.kr), J-Hoon Bae (baejh@etr.re.kr), Chan-Won Park (cwp@ etr.re.kr), and Hyo-Chan Bang (bangs@etr.re.kr) are wth the I Convergence echnology Research Laboratory, ERI, Daejeon, Rep. of Korea. Hyung Chul Park (correspondng author, hcpark@seoultech.ac.kr) s wth the Department of Electronc and I Meda Engneerng, Seoul Natonal Unversty of Scence and echnology, Seoul, Rep. of Korea. I. Introducton Rado-frequency dentfcaton (RFID), a non-contact automatc recognton technology, s a technology for recognzng electronc tags on products usng a rado frequency. RFID technology can be classfed, n a broad sense, nto a passve RFID system and an actve RFID system. In a passve RFID system, a tag s not suppled wth power from a battery; hence, t must communcate wth a reader based on backscatter by generatng self-power n response to a carrer sgnal from the reader. A passve RFID system can be used for varous applcatons n comparson to barcodes because t can provde nformaton on ndvdual objects wthout requrng the use of batteres n tags. here has been a sgnfcant amount of research on mprovng passve RFID system technologes. Prevous work [] [2] has manly focused on developng an effcent algorthm to enhance nventory effcency. In [3] [5], many knds of recever structures for demodulatng backscattered tag sgnals were ntroduced. In addton, wth the remarkable development n low-power semconductor technologes, nexpensve RFID tags are becomng a realty [6] [7]. However, most research has focused on demodulaton, protocol algorthms for readers, and mplementaton ssues related to tags. hese days, hgh-memory passve RFID tags are requred for applcatons that need to store data beyond an dentfcaton number. For example, manufacturng engneers for the aerospace and car ndustres have seen benefts n storng nspecton, brth, and repar records. In addton, accordng to an ncrease n the amount of data handled, fle management and securty servces for RFID are requred [8] [9]. Nevertheless, exstng passve RFID systems have problems n terms of performance and transmsson speed. radtonal ERI Journal, Volume 37, Number 2, Aprl 25 25 Sang-Hyun Mo et al. 24 http://dx.do.org/.428/etrj.5.234.

passve RFID tags wth a Mller-modulated subcarrer based on the ISO/IEC 8-63 nternatonal standard [] have data rates of up to 32 kbps, whch s nsuffcent for readng a large amount of data n a tag memory. For ths reason, there s a necessty to develop new modulaton technques for ncreasng the data rate of RFID tags. In some prevous studes, hgh data rate RFID systems have been studed [] [3]. In [] and [2], M-ary quadrature ampltude modulated (M-QAM) backscatter was proposed. Wth M-QAM backscatter, tags can transmt log 2 M data bts per symbol perod; however, tags can acheve only a 3 BER performance wth E b /N between 9 db and db. here also exsts an mplementaton ssue regardng the accuracy dfference between the deal and measured mpedance values. In [3], a dgtal RF-transmttng scheme wth a Mbps data rate s descrbed, but ths transmtter needs to be battery powered. o overcome ths weakness, we propose a new transmsson scheme to ncrease the data rate of a tag wthout degradng the demodulaton performance or requrng addtonal battery power. In ths paper, we present twodmensonal b-orthogonal sgnalng and orthogonal multplexng technques for passve RFID tags. he data rate of the proposed scheme s two to three tmes faster than a Mller modulated subcarrer wth the same occuped bandwdth. he READ command n the ISO/IEC 8-63 nternatonal standard [] allows a reader to read part of or all of a tag memory. When WordCount equals zero n the READ command, a tag should reply the contents of the chosen memory bank startng at WordPtr and endng at the end of the bank. However, accordng to an ncrease n the read data sze of a tag memory, the probablty of the packet error s also ncreased. herefore, when a reader tres to read the contents n a hgh-memory tag, the reader should repeatedly send a READ command wth the proper WordCount untl obtanng all nformaton n the memory. For ths reason, the data rate of a reader s also mportant to ncrease the readng speed of the tag memory. We propose a new encodng scheme for reader-to-tag communcaton that can provde a two-fold faster data rate n comparson to pulsenterval encodng (PIE). It can contrbute to an mprovement n both the readng and nventory speeds. he rest of ths paper s organzed as follows. Secton II presents a new encodng scheme and decodng algorthm for forward lnk communcaton. In Secton III, the modulaton and demodulaton sgnal model for a hgh-speed Mller (HS- Mller) modulated subcarrer are descrbed. In Secton IV, we descrbe the transmtter and recever structures for the proposed transmsson scheme based on MALAB/Smulnk, as well as provdng the expermental results. In Secton V, some concludng remarks are provded. II. Proposed Extended Pulse-Interval Encodng (E- PIE) Scheme. Data Encodng o ncrease the data rate of a reader, a reader encodes the reader command usng E-PIE. E-PIE uses four knds of symbol waveforms and can transmt two bts per symbol waveform. In Fg., E-PIE symbol waveforms are descrbed. E-PIE symbol waveforms are composed of a sgnal havng a dfferent length wth hgh and low values. Hgh values represent the transmtted CW, and low values represent the attenuated CW. he ar s the basc tme reference unt for reader-to-tag sgnalng, and here, t represents the duraton of symbol- and symbol-. he reader can communcate usng ar values wthn the range of 6.25 μs to 25 μs. A.5 ar value s used as the duraton of symbol-2 and symbol-3. he conventonal PIE scheme encodes one bt per symbol waveform based on the nformaton of dfferent symbol duratons. he symbol duraton of data- s ar, and the symbol duraton of data- s between.5 ar and 2 ar. However, the E-PIE scheme encodes two bts per symbol waveform based on two dfferent lengths of hgh and low values. For example, the frst data bt determnes the total symbol duraton between ar and.5 ar, and the second data bt determnes the length of the low value between.5 ar and.265 ar. RF parameters for generatng E-PIE symbol waveforms follow the current ISO/IEC 8-63 standard []; thus, when we apply the E-PIE scheme, there wll be no problems wth tag operatons. 2. Data Decodng heoretcally, both PIE and E-PIE have almost an error-free Symbol- (bt- bt-) Symbol- (bt- bt-) Symbol-2 (bt- bt-) Symbol-3 (bt- bt-) ar.5 ar.265 ar.5 ar.5 ar.265 ar Fg.. Proposed extended pulse-nterval encodng (E-PIE) symbol waveforms. 242 Sang-Hyun Mo et al. ERI Journal, Volume 37, Number 2, Aprl 25 http://dx.do.org/.428/etrj.5.234.

Yes Sym- (bt- bt-) CLK (.92 MHz) Yes Measure low nterval low low >(.5ar+.265ar)/2 Low nterval count E-PIE symbol duraton count Start Measure symbol duraton s No Yes ar s<(ar+.5ar)/2.5 ar Fg. 2. Decodng algorthm for E-PIE symbols. No Measure low nterval low low >(.5ar+.265ar)/2 Sym- Sym- Sym-2 Sym-3 No d, d, d 2, Data HS-Mller encodng d, d 2, HS-Mller encodng d, d, d 2, d 3, SW wth Data Demultplexer LF HS-Mller d, d 3, encodng SW wth LF2 ASK modulaton SW wth LF cos(2πf c t) (a) (b) ASK modulaton ASK modulaton cos(2πf c t) cos(2πf c t) x (t) x (t) x 2 (t) Fg. 3. Block dagram of a tag modulator based on an HS-Mller subcarrer: (a) sngle HS-Mller subcarrer and (b) double HS-Mller subcarrers. demodulaton performance because the mportant factor for a tag operaton s not the senstvty of the tag but the receved power from the reader. As long as the tag can wake up, t can have a suffcent sgnal-to-nose rato (SNR) value (that s, almost db) and therefore there wll be no bt error when a tag demodulates E-PIE symbols. o decode the E-PIE sgnal, only one more counter s requred to measure the low nterval length, and a low-clock rate can be used. A decodng algorthm for E-PIE symbols s descrbed n Fg. 2. Frst, a tag needs to measure the symbol duraton. A tag nterprets the receved E-PIE symbol as symbol- (Sym-) and symbol- (Sym-) when the measured symbol duraton s shorter than the average symbol duraton. After that, the tag agan compares the measured low nterval length wth the average low nterval length. When the measured low nterval length s longer than the average low nterval length, the tag fnally determnes the receved E-PIE symbol to be symbol-. A decodng algorthm for other E-PIE symbols s smlar to symbol-. III. Proposed HS-Mller Encodng Scheme. Modulaton of HS-Mller Subcarrer Sgnals A tag communcates wth a reader usng backscatter modulaton n whch the tag swtches the reflecton coeffcent n accordance wth the data sgnal. A tag encodes the backscattered data as the HS-Mller of a subcarrer. For hgh-speed communcaton, a tag uses up to two transmt antennas. When a tag transmts the backscattered data usng two antennas, a tag generates double HS-Mller subcarrers, where each subcarrer has a dfferent lnk frequency (LF). A block dagram of a tag modulator s shown n Fg. 3. he backscattered bts are converted nto 4-ary symbols, and the symbols are encoded usng the symbol waveforms n the 4-ary b-orthogonal sgnal set. hs sgnal set wll be explaned later. For generatng the HS-Mller subcarrer sgnal, the output waveform of the HS- Mller encoder s multpled by a square wave (SW) wth the defned LF, whch s M-tmes the symbol rate, /. herefore, an HS-Mller subcarrer sgnal shall contan exactly M subcarrer cycles per symbol duraton,. he HS-Mller subcarrer sgnal s modulated by an ASK modulator and upconverted usng a receved carrer sgnal from the reader. he transmtted sgnal, x (t), of the th tag antenna s gven by x () t s ( tm) w ()cos(2π t f t), (), j c m where m s the mth order of the HS-Mller symbols, s, j() t s the jth symbol waveform n the 4-ary b-orthogonal sgnal set for the th HS-Mller subcarrer ( j {,2,3,4}), w (t) s an SW for the th HS-Mller subcarrer, and f c s the carrer frequency. he N b-orthogonal sgnal waveform set can be obtaned from an orgnal orthogonal set of N/2 sgnals by augmentng t wth the negatve of each sgnal. herefore, the N sgnal waveforms are represented as the followng set: s () t s (), t s (),..., t s (), t s (), t s (),..., t s () t.(2) j 2 N/2 2 N/2 he orgnal orthogonal waveforms have to satsfy the followng orthogonal condton: Es f k l, s () ()d k t sl t t (3) f k l, where k and l N, 2,..., / 2. Such a set of b-orthogonal waveforms can be represented as a set of N/2 dmensonal orthogonal vectors. hat s, ERI Journal, Volume 37, Number 2, Aprl 25 Sang-Hyun Mo et al. 243 http://dx.do.org/.428/etrj.5.234.

E s 2 (t) (a) s 2 E s s (t) s 3 = s Fg. 4. Sgnal constellatons for Mller and HS-Mller sgnals: (a) orthogonal sgnal for Mller and (b) 4-ary borthogonal sgnal for HS-Mller. E s 2 (t) (b) s 2 E s s 4 = s 2 s (t) Symbol- (bt- bt-) Symbol-2 (bt- bt-) Symbol- (bt- bt-) Symbol-3 (bt- bt-) Symbol-4 (bt- bt-) Symbol-2 (bt- bt-) (a) M = 2 Symbol- (bt- bt-) Symbol-2 (bt- bt-) Symbol-3 (bt- bt-) Ampltude Ampltude s (t) s 3 (t) me (t) me (t) Ampltude Ampltude Fg. 5. Bass waveforms for HS-Mller encodng. s 2 (t) /2 s 4 (t) s ( E,,,..., ),, s (,,,..., E ), s N/2 s /2 me (t) me (t) s ( E,,,..., ),, s (,,,..., E ). ( N/2) s N s (4) Fgure 4 llustrates the sgnal constellatons correspondng to N = 4 b-orthogonal sgnals for HS-Mller encodng and twodmensonal orthogonal sgnals for Mller encodng. In ths paragraph, HS-Mller encodng and a method for generatng an HS-Mller subcarrer are explaned. In the HS- Mller encodng process, every two bts are mapped to HS- Mller symbols, and HS-Mller symbols are encoded usng 4-ary b-orthogonal waveforms. In comparson to Mller encodng (M = 4), HS-Mller encodng can transmt data twotmes faster than Mller encodng durng the same symbol perod. For generatng an HS-Mller subcarrer, an HS-Mller encoded waveform s multpled by SW at M-tmes the symbol rate. Fgure 5 shows b-orthogonal waveforms for HS-Mller encodng, and HS-Mller subcarrer sgnals are descrbed n Fg. 6. When a tag transmts backscattered data usng orthogonal double subcarrers, t can acheve a much faster data rate. o satsfy the orthogonal characterstc between two subcarrers, the LF of one HS-Mller subcarrer s L-tmes faster than the other subcarrer. herefore, usng an orthogonal multplexng technque, there s no performance degradaton at the reader. Symbol-3 (bt- bt-) Symbol-4 (bt- bt-) Symbol-4 (bt- bt-) (b) M = 4 (c) M = 8 Fg. 6. HS-Mller subcarrer sgnals: (a) M = 2, (b) M = 4, and (c) M = 8. able. Data rates of Mller subcarrer and HS-Mller subcarrer. Encodng type M Assumed LF (khz) Data rate (kbps) 2 64 64 Sngle HS-Mller 4 64 32 8 64 6 2 64/32 96 Double HS-Mller 4 64/32 48 8 64/32 24 2 64 32 Mller 4 64 6 8 64 8 In able, we brefly compare the data rate of an HS-Mller subcarrer wth that of a Mller subcarrer. he data rate of each HS-Mller subcarrer can be calculated as Data rate (LF/ M ) 2. (5) In the followng secton, we descrbe the demodulaton structures and methods for HS-Mller subcarrer sgnals. 2. Demodulaton of HS-Mller Subcarrer Sgnals For demodulaton of an HS-Mller subcarrer sgnal, the maxmum a posteror probablty (MAP) detector s requred. Assume that an HS-Mller subcarrer sgnal s used to transmt backscattered data through an AWGN channel, and wthout 244 Sang-Hyun Mo et al. ERI Journal, Volume 37, Number 2, Aprl 25 http://dx.do.org/.428/etrj.5.234.

consderaton of a low-pass flter and other RF component effects, the receved sgnal, r(t), s smply defned by rt () s() twt () nt () for t, (6) j where n(t) s whte Gaussan nose wth a power spectrum of N /2 watts/hertz, s j (t) s the jth symbol waveform n the N borthogonal sgnal set ( j {,2,..., N} ), and w(t) s an SW for generatng an HS-Mller subcarrer. he optmum MAP detector can be mplemented usng sgnal correlators for an orgnal orthogonal set of N/2 sgnals. he receved sgnal r(t) s cross-correlated wth each of the N/2 reference orthogonal sgnal waveforms, and correlator outputs are sampled at a samplng rate of /. he sampled sgnal r s gven by where r r() t s () t w()d t t for,2,..., N /2 En f j, En f j N / 2, n else, (7) () () ()d for,2,..., /2 (8) n n t s t w t t N and E s the symbol energy for each sgnal waveform. An example of the optmum MAP detector for an HS-Mller subcarrer sgnal s descrbed n Fg. 7. he optmum MAP detector observes four correlator outputs and chooses the one wth the largest magntude. Next, the detector determnes the HS-Mller symbols accordng to the sgn of the chosen correlator output. he decson rule can be wrtten as sˆ arg max{ r } and r, (9) sˆ arg max{ r } and r (,2,..., N / 2). ( N/2) We compare the probablty of a bt error occurrng between the Mller subcarrer and HS-Mller subcarrer schemes. he Mller subcarrer scheme uses two-dmensonal orthogonal waveforms to encode the transmt data. In two-dmensonal orthogonal sgnalng, the probablty of a bt error can be expressed as r(t) s (t)w(t) s 2 (t)w(t) k ( k) k ( k) ()d t ()d t t=k t=k Fg. 7. Optmum MAP detector structure for HS-Mller subcarrer sgnal. r,k r 2,k Detector Output decson 2 x 2 Eb/ N 2 Pe,Mller Q( x) e dx 2π () Q Eb / N, where E b s bt energy. Now, we consder the probablty of a bt error for an HS- Mller subcarrer sgnal. he condtonal probablty of a correct decson for equally lkely messages s as follows. We assume that s (t) s transmtted. pr ( s(), t r) Pr( rall rr: j s(), t r) j N/2 {Pr( r all rjr s( t), r )} N/2 r/ N/2 2 r /2 e d r. r / N /2 () Averagng over the probablty densty functon of r, we can derve the correct decson probablty as r/ N/2 2 r /2 pr s( t) e dr r e π 2N 2 ( re) / N N/2 d r, (2) where N 2 s the varance of the Gaussan nose n, and E s the symbol energy for the s (t) waveform. For an HS-Mller subcarrer scheme, gven N = 4, the probablty of a bt error can be wrtten as e,hs-mller b P Q 2 E / N. (3) hus, the HS-Mller subcarrer scheme acheves an SNR that s about 3 db better than that of the Mller subcarrer scheme for the same gven BER. Moreover, bass waveforms for FM encodng are the same as Mller encodng. HS-Mller can have a demodulaton performance that s about 3 db better than that of FM encodng. IV. Desgn of Smulaton Model and Expermental Results. MALAB/Smulnk Model for ag-to-reader Communcaton Based on HS-Mller Subcarrer In ths secton, we present a MALAB/Smulnk model for the modulaton and demodulaton of HS-Mller subcarrer sgnals. hs smulaton model can be used for examnng the characterstcs of the transmtted tag sgnal and the demodulaton performance of the recever structure. Fgure 8 shows a top-level block dagram of the tag-to-reader communcaton model based on an HS-Mller subcarrer. he smulaton model conssts of a transmtter, an AWGN channel, ERI Journal, Volume 37, Number 2, Aprl 25 Sang-Hyun Mo et al. 245 http://dx.do.org/.428/etrj.5.234.

Hgh-speed RFID tag model Subcarrer Subcarrer 2 Sg.In Sg.Out No nose AWGN channel cos cos Mxer Mxer 2 + + Sg.In Carrer RF x/rx I-ch Sg. RF x/rx Q-ch Sg. Reader RF model [RX_CH_I] [RX_CH_Q] 4-HS-Mller-data Repeat 32 8-HS-Mller-data ag modulator Zero-order hold B-FF Spectrum scope Spectrum [RX_CH_I] [RX_CH_Q] I-ch A/D Q-ch A/D I-ch Sg. Q-ch Sg. I-ch Sg. ln Q-ch Sg. ln Decmaton FL I-ch Sg. Out Q-ch Sg. Out I-ch Sg. FL Out Q-ch Sg. FL Out Correlator x Rx Error rate calculaton BER calculaton BER dsplay BER # bt errors # bts LF for M = 4 2 for M = 8 LF 3-stage decmaton flter Correlator Fg. 8. op-level block dagram of the tag-to-reader MALAB/Smulnk model based on HS-Mller subcarrer. Magntude-squared (db) Magntude-squared (db) 2 3 4 5 6 7 8 9 2 3 4 5 6 7 8 9..8.6.4.2.2.4.6.8. Frequency (MHz) (a)..8.6.4.2.2.4.6.8. Frequency (MHz) (b) Fg. 9. Frequency response of the transmtted sgnal: (a) Mller subcarrer (M = 4, LF = 3 khz) and (b) HS-Mller subcarrer (M = 4, LF = 3 khz). and a recever. he transmtter block ncludes a bt data generator, an HS-Mller encoder, and an HS-Mller subcarrer generator and s able to select parameter M, the number of subcarrers, and the LF of each subcarrer. Fgure 9 shows the frequency spectrums of a Mller subcarrer sgnal and an HS-Mller subcarrer sgnal. In partcular, the frequency spectrum of an HS- Mller subcarrer sgnal does not have a dc component; thus, the recever s not affected by the dc-offset nose. s () ()d for,2,...,. twt t N (4) As shown n Fg. 9, the frequency response of an HS-Mller subcarrer s very smlar to the frequency response of a Mller subcarrer; therefore, the same analog and dgtal flters for Mller decodng can be used at the transmtter and recever. In addton, our proposed scheme can be ftted based on the regulaton of the channel sgnalng defned n the ISO/IEC 8-63 nternatonal standard [], and can be used for a dense reader envronment. he AWGN channel model can be set up for both the SNR value and the exstence of nose. After passng through the AWGN channel, the baseband sgnal of the HS-Mller subcarrer s up-converted by the receved carrer sgnal from the reader. At the recever block, the RF module down-converts the receved RF sgnal. he down-converted sgnal s fltered by a low-pass flter and over-sampled by a dgtal-to-analog converter. he baseband modem module 246 Sang-Hyun Mo et al. ERI Journal, Volume 37, Number 2, Aprl 25 http://dx.do.org/.428/etrj.5.234.

able 2. Smulaton parameters for double subcarrers. (a) Parameter Encodng LF (khz) Descrpton Double HS-Mller Double subcarrer: LF 3, LF2 6 Double subcarrer wth nterference: LF2 45, LF2 6 (b) 2 Fg.. Input and recovered streams of the HS-Mller symbols: (a) source HS-Mller symbols and (b) recovered HS- Mller symbols. BER 3 4 conssts of a decmaton flter and a correlator for an HS-Mller symbol decson. A decmaton flter s used for down-samplng the receved baseband sgnal accordng to the LF and samplng frequency. he correlator calculates the cross-correlaton energy of both the receved baseband sgnal and the reference HS-Mller subcarrer sgnal and provdes the output to the HS- Mller symbol decoder. As we explaned n Secton III, the HS- Mller symbol decoder compares the correlator outputs and determnes the HS-Mller symbol. he outputs from both the random bt generator block and the HS-Mller decodng block are smultaneously sent to the error rate calculaton block, and the BER s computed. As shown n Fg., the HS-Mller symbols are well recovered from the transmtted HS-Mller symbol wth AWGN usng the correlator and symbol decoder n our desgned smulaton model. Compared wth the HS- Mller source symbols, whch have four knds of quantzaton levels, the recovered HS-Mller symbols smply have a tme delay owng to several flterng processes. Our tag-to-reader lnk layer smulaton model provdes the facltes to smulate and test a passve RFID system at a hgh level of abstracton. In the followng secton, we evaluate the performance of the proposed scheme usng our MALAB/Smulnk model. 2. Expermental Results In ths secton, we provde expermental results demonstratng the effectveness of our proposed transmsson scheme for hgh-speed tag-to-reader communcaton. We compare the BER performance of a sngle HS-Mller subcarrer scheme to both the FM and the Mller subcarrer scheme. In addton, we 5 6 Sngle HS-Mller subcarrer Double HS-Mller subcarrer Double HS-Mller subcarrer wth nterference FM/Mller subcarrer 2 3 4 5 6 7 8 9 E b /N (db) Fg.. BER comparson between HS-Mller subcarrer scheme and FM/Mller subcarrer scheme. evaluate the demodulaton performance for a double HS-Mller subcarrer scheme under specfc LF condtons n whch the LF of one subcarrer s twce as fast as that of the other subcarrer. Fgure shows the BER performance of the HS- Mller subcarrer scheme n the presence of AWGN. We compare the BER of the sngle HS-Mller subcarrer scheme to the theoretcal BER of both the FM and the Mller subcarrer. he smulaton results show that the sngle HS-Mller subcarrer scheme can acheve about a 3 db hgher BER performance than the Mller subcarrer at the same E b /N and that t has a BER of 5 at an E b /N of db. We also evaluated the demodulaton performance for the double HS- Mller subcarrer scheme. o demonstrate the advantage of orthogonal multplexng technques, we assume two types of LFs for double subcarrers. In able 2, the LFs of the double HS-Mller subcarrers are defned. In Fg., a BER comparson between the sngle HS-Mller and double HS- Mller subcarrers schemes s shown. he smulaton results show that the double HS-Mller subcarrer provdes the same performance as the sngle HS-Mller subcarrer when LF2 s twce as fast as LF. Under ths condton, these subcarrers are orthogonal; thus, there s no performance degradaton at the recever caused by mutual nterference. he orthogonal multplexng condton for the double HS-Mller subcarrer ERI Journal, Volume 37, Number 2, Aprl 25 Sang-Hyun Mo et al. 247 http://dx.do.org/.428/etrj.5.234.

BER 2 3 4 5 6 FM Mller subcarrer (M=2) Mller subcarrer (M=4) HS-Mller subcarrer (M=2) HS-Mller subcarrer (M=4) 2 3 4 5 6 7 8 9 E s /N (db) Fg. 2. BER comparson between HS-Mller subcarrer scheme and FM/Mller subcarrer scheme from the pont of vew of requred E s /N. scheme can be wrtten as LF/ LF2 M or LF2/ LF N, (5) where M and N are ntegers. When HS-Mller subcarrers are transmtted wth LF at 6 khz and LF2 at 45 khz, about a db performance degradaton s caused by mutual nterference. o reject such nterference, an nterference rejecton flter s requred at the recever. he data rate of FM encodng n ISO/IEC 8-63 [] s up to 64 Kbps, and the data rate of the Mller subcarrer s sgnfcantly slower than that of FM by a factor of two, four, or eght. However, a Mller subcarrer s wdely used n realfeld applcatons because a Mller subcarrer requres less E s /N for the gven BER and s more sutable for a dense reader envronment. o overcome the slow data rate of the Mller subcarrer, the proposed HS-Mller subcarrer s a hghly useful soluton for ncreasng the data rate of a tag wthout degradng the demodulaton performance. In Fg. 2, the HS-Mller subcarrer has the same demodulaton performance as the Mller subcarrer n terms of the requred E s /N ; however, t can acheve a two- to three-fold faster data rate than the Mller subcarrer. V. Concluson In ths paper, we presented new encodng methods sutable for hgh-speed UHF RFID communcatons. In forward lnk communcaton, a reader encodes the transmtted data usng E- PIE. E-PIE can acheve a two-fold faster data rate than PIE, and only one more counter s requred to decode E-PIE symbols at the tag. In backward lnk communcaton, the desgned subcarrer sgnals are modulated by M-state b- orthogonal bass waveforms and are transmtted usng up to two antennas. he proposed HS-Mller subcarrer scheme can mprove the data rates remarkably n comparson to a Mller subcarrer. For example, a sngle HS-Mller subcarrer scheme can acheve a two-fold faster data rate than a Mller subcarrer when the number of subcarrer cycles per symbol s four. Moreover, the frequency spectrum of HS-Mller subcarrer sgnals s very smlar to that of Mller subcarrer sgnals; thus, HS-Mller subcarrer sgnals satsfy the channel sgnalng regulatons defned n the ISO/IEC 8-63 standard. In addton, our proposed scheme s easy to mplement. he tag complexty of a sngle HS-Mller subcarrer scheme s the same as that of a Mller subcarrer; however, when a tag uses double HS-Mller subcarrers, the RF block of the tag complexty s ncreased owng to the use of two load modulators. he reader complexty s the same n sngle subcarrer transmsson, and the reader complexty of double HS-Mller subcarrers s slghtly ncreased because the sgnal detecton and synchronzaton blocks are the same as n a Mller subcarrer, although two demodulaton blocks are requred for double HS-Mller subcarrers. o demonstrate our proposed scheme, we developed a smulaton model for hghspeed backward lnk communcaton based on HS-Mller subcarrers. We utlzed ths smulaton model usng the MALAB/ Smulnk tool and smulated the BER performance n an AWGN channel. For the sngle HS-Mller subcarrer scheme, the smulaton results ndcate that a BER performance of 5 can be acheved at an E b /N of db and 9 db, respectvely. In addton, n double HS-Mller subcarrer transmsson mode, the lnk frequences for each subcarrer have to satsfy the orthogonal multplexng condton, whch s defned n (5), to acheve a faster data rate than a sngle subcarrer wthout degradng the performance. Lkewse, the proposed scheme sgnfcantly mproves the data rate, spectral effcency, and BER performance n comparson to the Mller subcarrer scheme. Addtonal future work ncludes constructng a tag emulator that mplements an HS-Mller subcarrer, and evaluatng the real-world performance n a passve RFID envronment. References [] L. Kang et al., DDC: A Novel Scheme to Drectly Decode the Collsons n UHF RFID Systems, IEEE rans. Parallel Dstrb. Syst., vol. 23, no. 2, Feb. 22, pp. 263 27. [2] C. Jn et al., An Effcent Collson Detecton Scheme for Generaton-2 RFID Systems, Int. J. Comput. Sc. Issues, vol. 9, no., Sept. 22, pp. 29 39. [3] J.-H. Bae et al., Desgn of Reader Baseband Recever Structure for Demodulatng Backscattered ag Sgnal n a Passve RFID 248 Sang-Hyun Mo et al. ERI Journal, Volume 37, Number 2, Aprl 25 http://dx.do.org/.428/etrj.5.234.

Envronment, ERI J., vol. 34, no. 2, Apr. 22, pp. 47 58. [4] I. Mayordomo et al., Desgn and Implementaton of a Long- Range RFID Reader for Passve ransponders, IEEE rans. Mcrow. heory echn., vol. 57, no. 5, May 29, pp. 283 29. [5] S.-C. Jung et al., A Reconfgurable Carrer Leakage Canceler for UHF RFID Reader Front-Ends, IEEE rans. Crcuts Syst., vol. 58, no., Jan. 2, pp. 7 76. [6] V. Plla et al., An Ultra-Low-Power Long Range Battery/Passve RFID ag for UHF and Mcrowave Bands wth a Current Consumpton of 7 na at.5 V, IEEE rans. Crcuts Syst., vol. 54, no. 7, July 27, pp. 5 52. [7] H. Nakamoto et al., Passve UHF RF Identfcaton CMOS ag IC Usng Ferroelectrc RAM n.35 μm echnology, IEEE J. Sold-State Crcuts, vol. 42, no., Jan. 27, pp.. [8] S. Cho et al., A Fully Integrated CMOS Securty-Enhanced Passve RFID ag, ERI J., vol. 36, no., Feb. 24, pp. 4 5. [9] Y.S. Kang, D. Cho, and D.-J. Park, Comments on an Improved RFID Securty Protocol for ISO/IEC WD 2967 6, ERI J., vol. 35, no., Feb. 23, pp. 7 72. [] ISO/IEC Std. 8-63, Informaton echnology - Rado Frequency Identfcaton for Item Management - Part 63: Parameters for Ar Interface Communcaton at 86 MHz to 96 MHz ype C, ISO/IEC JC/SC 3, 23. [] S.J. homas and M.S. Reynolds, QAM Backscatter for Passve UHF RFID ags, IEEE Int. Conf. RFID, Orlando, FL, USA, Apr. 4 6, 2, pp. 2 24. [2] S.J. homas and M.S. Reynolds, A 96 Mbt/sec, 5.5 pj/bt 6 QAM Modulator for UHF Backscatter Communcaton, IEEE Int. Conf. RFID, Orlando, FL, USA, Apr. 3 5, 22, pp. 85 9. [3] K. Suzuk, M. Ugajn, and M. Harada, A Mbps.6 μa Mcro- Power Actve-RFID CMOS LSI for 3 MHz Frequency Band, IEEE/M-S Int. Mcrow. Symp., Honolulu, HI, USA, June 3 8, 27, pp. 57 574. Sang-Hyun Mo receved hs BS and MS degrees n electronc and electrcal engneerng from Pohang Unversty of Scence and echnology (POSECH), Rep. of Korea, n 26 and 28, respectvely. Snce 28, he has been wth the Electroncs and elecommuncatons Research Insttute, Daejeon, Rep. of Korea, as a researcher. Hs current research nterests nclude RFID systems, wreless communcaton systems, and modem desgns for communcaton systems. J-Hoon Bae receved hs BS degree n electronc engneerng from Kyungpook Natonal Unversty, Daegu, Rep. of Korea, n 2 and hs MS degree n electrcal engneerng from Pohang Unversty of Scence and echnology (POSECH), Rep. of Korea, n 22. Snce 22, he has been wth the Electroncs and elecommuncatons Research Insttute, Daejeon, Rep. of Korea, as a senor researcher. In addton, snce 23, he has been pursung hs PhD degree n radar sgnal processng at POSECH. Hs current research nterests nclude modulaton/demodulaton of RFID readers, RFID systems, radar sgnal processng, array antennas, and optmzaton technques. Chan-Won Park receved hs BS and MS degrees n computer engneerng from Kwangwoon Unversty, Seoul, Rep. of Korea, n 993 and 996, respectvely. From 996 to 999, he was a member of the ASIC engneerng staff, KAIS IDEC, Daejeon, Rep. of Korea. Snce 999, he has been workng at the Electroncs and elecommuncatons Research Insttute (ERI), Daejeon, Rep. of Korea and has been a drector of the Smart hngs Cognton Research Secton at ERI snce 2. Hs research nterests nclude wreless LAN, Io, M2M, RFID, and system of chps. Hyo-Chan Bang receved hs BS and MS degrees n management engneerng and ndustral engneerng from the Hokkado Insttute of echnology, Japan, n 995 and 997, respectvely. In 27, he completed a doctoral course n computer engneerng from Chungnam Natonal Unversty, Daejeon, Rep. of Korea. From 997 to 999, he was an assocate researcher at K, Seoul, Rep. of Korea. In 2, he joned the Electroncs and elecommuncatons Research Insttute, Daejeon, Rep. of Korea, where he s currently a managng drector of the Io Convergence Research Department. Hs current research nterests nclude platform, network, and devce technologes related to Io; dstrbuted systems; moble computng; and nformaton securty. Hyung Chul Park receved hs BS, MS, and PhD degrees n electrcal engneerng from the Korea Advanced Insttute of Scence and echnology, Daejeon, Rep. of Korea, n 996, 998, and 23, respectvely. From 23 to 25, he was an SoC desgn engneer wth Hynx Semconductor, Seoul, Rep. of Korea. From 25 to 2, he was an assstant professor at Hanbat Natonal Unversty, Daejeon, Rep. of Korea. In 2, he joned the faculty of ERI Journal, Volume 37, Number 2, Aprl 25 Sang-Hyun Mo et al. 249 http://dx.do.org/.428/etrj.5.234.

the Department of Electronc and I Meda Engneerng, Seoul Natonal Unversty of Scence and echnology, Rep. of Korea, where he s currently an assocate professor. Hs current research nterests nclude wreless modulaton/demodulaton algorthms, system desgn/mplementaton, and nterface study between RF/IF stages and dgtal sgnal processng. 25 Sang-Hyun Mo et al. ERI Journal, Volume 37, Number 2, Aprl 25 http://dx.do.org/.428/etrj.5.234.