RFID DESIGN STUDIES Dr. KVS Rao Prof. Raj Mittra Intermec Technologies Everett, WA, USA Prof. Raj Mittra Pennsylvania State University University Park, PA, USA.

Applications Electronic Toll Collection Access Control Animal Tracking Inventory Control Tracking Runners in Races!

Radio Frequency Identification (RFID) System - Background Information Introduction Radio Frequency Identification (RFID) System - Background Information Applications of RFID High Frequency (13.56MHz) Supply Chain Wireless Payment Libraries Book ID Ultra High Frequency (902 – 928 MHz) Sensors Libraries Microwave Frequency (2.45GHz) Electronic Toll Payments

Design Challenges Small Size Planar UHF Frequency Allocation Europe 866-869 MHz North America 902-928 MHz Impedance Matching ASIC Chip: High Capacitive Value, Small Resistive Value Environmental Conditions

Antenna Parameters Characteristic Impedance Power 3-Dimensional Radiation Patterns Maximum Directivity where Za = Ra + j Xa is the antenna impedance, and Zs = Rs + j Xs is the source impedance. where Umax is the radiation intensity in maximum direction, and Prad is the total radiated power.

Hybrid loop antenna Top View of the antenna Length of the antenna ~ one operating wavelength in free space Outer loop terminated by inner loop  size reduction Simple structure (one layer of dielectric substrate) Antenna impedance must be highly inductive

Hybrid loop antenna (cont‘d) Top View of the modified antenna Realize a high value for the inductance by: Changing the loop area (L ~ A) Changing the length of the perimeter

Hybrid loop antenna : First design Input Impedance Perimeter of the loop antenna : 244 mm Size used by the antenna : 39 x 40 (mm) Resonance frequency : ~0.5 GHz and 1.26 GHz Reactance +300 Ω at 1.01 GHz

Hybrid loop antenna : First design (cont‘d) Far Field Pattern (normalized) xy-plane xz- (blue) and yz-(red) plane Not omnidirectional Pattern in the xy-plane  Non strictly symmetry of the geometry

Hybrid loop antenna : Parametric study Changing the length of the inner loop Li Li = 16 (red graph) mm  Li = 10 mm (blue graph) Perimeter  , Loop area  : L  Perimeter 244 mm  232 mm

Hybrid loop antenna : Parametric study Changing the width of the inner loop Wi Wi = 28 (red graph) mm  Wi = 30 mm (blue graph) Perimeter  , Loop area  : L  Perimeter 248 mm  252 mm

Hybrid loop antenna (cont‘d) Far Field Pattern and Current distribution at 910 MHz Current distribution : small current in the top part of the antenna  small influence on the inductance  Meandering

Meandered Hybrid loop antenna Top View of the meanderd antenna Perimeter 252 mm  302 mm Maximum percentage at 910 MHz

Meandering technique reduces the size of the antenna Hybrid Loop Antenna OBSERVATIONS Length of the antenna has a greater effect on the input impedance more than does the loop area Meandering technique reduces the size of the antenna Small percentage power delivered to the antenna  attributable to very small resistive part of the input impedance The developed design did not prove to be too useful

Top and Side Views of the Antenna Structure Dual cross-dipole Top and Side Views of the Antenna Structure Cross plarization sens can als be greatly improve by using more than one indep.antenna on the tag . Dual diple – provides operation that is almost orientation independent Meandering dipole  size reduction Cross-polarization sensitivity   dual dipole Ground plane  can act as reflector  gain 

Dual cross-dipole : Design#1 Input Impedance Length of the antenna : 218 mm ~0.66 λ (at 910 MHz) Area used by the antenna : 51 x 51 (mm) Reactance is too small in the desired frequency  Length of the antenna  Resistive part is again very small

Dual cross-dipole : Design#2 Top View “Load bar“ is added Length of the antenna : 258 mm ~ 0.78 λ (at 910 MHz) f300 at 900 MHz

Dual cross-dipole : Design#2 Far Field Pattern/Power delivered to the antenna/ Axial Ratio ~80 % of the power is delivered to the antenna Narrow bandwidth (10.5 MHz more than 50 % is delivered) Min. AR  3 dB (860 MHz – 960 MHz)

Dual cross-dipole : Parametric study Influence of the height of the antenna Decreasing h, increases the resonance frequency By varying the height, input impedance can be adjusted for a good matching

Dual cross-dipole : Parametric study Influence of the dielectric constant Input Impedance and new design Increasing the dielectric constant , drops the resonance frequency  length of the antenna  Area used by the antenna was decreased ~ 19 % by using a higher dielectric (4 instead of 2.2 ) Max. Power delivered to the antenna was sligthly higher for the case with the higher dielectric constant (79 % vs 86 % ) Bandwidth wasn‘t influenced

Dual cross-dipole : New design New design/Power delivered to the antenna Area used by the antenna reduced ~ 35 % compared to the inital design (second case) and ~21 % compared to the previous case Max percentage for the power plot : ~81 % (79 % second case / 86 % previous case) Bandwidth didn‘t change

Inductively coupled Feed Top View and structure The proposed feed structure as well as the structure of the antenna system is shown in Fig. The antenna is composed of a small rectangular loop and a radiation body, which are coupled inductively. The radiation body correspond to the antenna D3) explored in the previous chapter. The feeding loop and the antenna are separated by a substrate with the substrate εr and the height h2 .   The strength of the inductive coupling is controlled by the distance between the feed and the antenna body (h2), as well as by the area of the loop (shape) .Fig shows the equivalent circuit of the inductively coupled feed structure. The inductive coupling is modelled by a transformer Strength of the coupling depends on h2 and the size of the loop Inductive coupling modeled by a transformer Analyzing the input impedance by varying the size and shape of the loop

Inductively coupled Feed (cont‘d) Changing length of the loop There are two range of frequency where the antenna can operate, before and after the inductance gets his maximum. To get the whole system in the right range, the antenna size could be reduced. But then we will have a very bad percentage of the power, which is transferred to the antenna, because of the very small resistive part in this frequency range Increasing the loop size, increase the inductance With this method the reactance increases ~200 Ω Two operating range frequency Antenna size needs to be adjusted (increased)

Inductively coupled Feed (cont‘d) Changing the shape of the feeding loop Same experiment as before (changing the size of the loop) For one design we realized a very high percentage of power delivered (98 % at 899 MHz) Bandwidth was narrow

Inductively coupled Feed (cont‘d) OBSERVATIONS Narrow bandwidth Operating frequency can be varied by changing the size of the feeding loop Antenna size must be increased to operate in the desired frequency range if we use a square loop.

Antenna Measurement Top View of the antenna

Input Impedance Comparison : Measurement et Simulation

Field Pattern normalized (910 MHz) - Comparison Measurement Simulation Anechoic chamber not ideal for 910 MHz Different feeding part (balun for measurement) Infinite substrate size used for simulation

PLATFORM-TOLERANT RFID DESIGNS

Dual-Band PIFA Design ASIC Chip: Zc=10-j160 [W] at 867 MHz Parameter Size (mm) L 62 W 51.3 H 3 S 5 gap (867/915 MHz) 1 gap (867/940 MHz) 1.9 er 2.35 ASIC Chip: Zc=10-j160 [W] at 867 MHz Zc=10-j150 [W] at 915 MHz Zc=10-j145 [W] at 940 MHz

Dual-Band PIFA Design Dual-band Frequency Operation Open-Ended Stub Gap Dimension and Stub Dimension Used to Tune Platform Tolerance Dominating Horizontal Current Distribution Widening Short, Vertical Inductance Reduced, Antenna Lowered

Dual-Band PIFA Design Mounting Materials Dimensions 900 mm x 900 mm Thickness=13 mm Cardboard (er=2.5) Glass(er=3.8) Plastic(er=4.7)

Impedance [867/915 MHz] Dual-Band PIFA Design

Power Dual-Band PIFA Design   Power (867 MHz) (915 MHz) (940 MHz) No Material 83.49 64.92 74.07 Cardboard 54.53 86.28 80.5 Amount Increased -28.96 21.36 6.43 Glass 54.81 80.72 72.9 -28.68 15.8 -1.17 Plastic 58.3 85.72 72 -25.19 20.8 -2.07

Radiation [867 MHz] Dual-Band PIFA Design No Material Cardboard Glass Plastic

Conclusions Dual-Band PIFA Design   Directivity (867 MHz) (915 MHz) (940 MHz) No Material 1.6841 1.832 1.8815 Cardboard 1.928 2.0704 2.3425 Amount Increased 0.2439 0.2384 0.461 Glass 2.4053 2.8135 3.9153 0.7212 0.9815 2.0338 Plastic 2.9936 3.4036 4.1411 1.3095 1.5716 2.2596

Environmental Change Dual-Band PIFA Design Cardboard Box 900 mm x 900 mm 4 x 4 Thickness=13 mm Metal sheet 450 mm x 450 mm 2 x 2 Height from Cardboard was Varied from 0 mm-20 mm Radiation [867 MHz] Metal 20 mm Under Cardboard No Metal

Environmental Change Dual-Band PIFA Design   Power (867 MHz) (915 MHz) Peak Directivity (867 MHz) Peak Directivity (915 MHz) No Metal 54.53 86.28 1.928 2.0704 Metal 0 mm 76.6 80.7 3.3105 3.0219 Amount Increased 22.07 -5.58 1.3825 0.9515 Metal 10 mm 91.08 73.11 3.1872 3.0167 36.55 -13.17 1.2592 0.9463 Metal 20 mm 73.25 76.82 3.2213 3.0599 18.72 -9.46 1.2933 0.9895

Ground Plane Optimization Dual-Band PIFA Design No Material Directivity (867 MHz) (915 MHz) Original GP 1.6841 1.832 1 Inch Larger 2.3265 2.3131 2 Inch Larger 2.9306 2.918 3 Inch Larger 3.6696 3.7427 10 Inch Larger 4.5087 4.4954   Cardboard 1.928 2.0704 2.6915 2.7059 3.2191 3.2618 3.1907 3.3583 4.6297 4.687 Glass Peak Directivity (867 MHz) Peak Directivity (915 MHz) Original GP 2.4053 2.8135 1 Inch Larger 2.5102 2.6094 2 Inch Larger 2.9178 3.0237 3 Inch Larger 2.7375 2.7357 10 Inch Larger 3.7178 4.1891   Plastic 2.9936 3.4036 2.9787 3.0035 3.0787 2.9965 3.1032 3.0408 3.0544 2.9356

Ground Plane Optimization Dual-Band PIFA Design

Inductively-Coupled Feed Loop PIFA Design Impedance Matching Inductively Coupled Feed Loop Gap dimension between loop and radiators is used to tune Designed to match Zc=10-j150 [W] at 915 MHz Platform Tolerance Reduced Current on Ground Plane

Inductively-Coupled Feed Loop PIFA Design Mounting Materials Dimensions 200 mm x 200 mm ( x ) Thickness=5 mm Cardboard (er=2.5) Glass with No Loss(er=3.8) Glass with Loss(er=2.5) and Loss Tangent 0.002

Optimization of Impedance in Free Space   Power 915 MHz [%] 940 MHz [%] Average Gap 9 mm 9.09 4.39 6.74 Gap 9.25 mm 86.09 41.45 63.77 Gap 9.5 mm 77.38 34.50 55.94 Gap 9.75 mm 15.28 13.00 14.14

Directivity & Radiation   Directivity (915 MHz) Directivity (940 MHz) No Mounting Material 3.47 3.3 Cardboard (er=2.5) 3.43 2.94 Amount Increased -0.04 -0.36 Glass No Loss (er=3.8) 3.4 3.25 -0.07 -0.05 Glass With Loss (er=2.5) and loss 0.002 3.36 --0.11 867 MHz No Material 867 MHz Cardboard

Optimization of Impedance for Cardboard

Impedance Inductively-Coupled Feed Loop PIFA

Impedance Inductively-Coupled Feed Loop PIFA

Power Before and After Optimization Inductively Coupled Feed Loop PIFA   Power (915 MHz) [%] Power (940 MHz) [%] Free Space 86.09 41.45 Cardboard 16.29 6.5 Cardboard Optimized 61.19 31.69 Glass 24.06 9.48 Glass Optimized 56.59 69.36 Glass with Loss 11.65 23.81 Glass with Loss Optimized 61.6 52.55

Performance Enhancement with Artificial Magnetic Conductors PMC Ground Image Currents In Phase with Original Currents PMC is reflective Low Profile Antennas High Impedance Surface Current is filtered at selected frequencies so tangential magnetic field is small while electric field is still large Suppression of Surface Waves=>Minimizes Backlobe Reflection Coefficient of +1 PEC Ground Reflects Half the Radiation Gain can be increased by 3 dB Image Currents Can Cancel Currents in Antenna Limitation on distance between ground and radiating elements (/4) Reflection Coefficient of -1

Metamaterials Challenges Electrical Ground Plane Redirect one-half of the radiation  gain can be increased by 3 dB Min. distance between antenna and ground : λ/4 Image currents cancel currents in antenna  poor radiation efficiency Metamaterials→Material that exhibit electromagnetic properties not found in nature EBG (Electromagnetic Band Gap) - Surface Subclass of metamaterials Can be designed to act as an AMC (Artificial Magnetic Conductor) ground plane In many antennas a flat metal sheet is used as a ground plane or reflector. The presence of a ground plane redirects one-half of the radiation into the opposite direction, improving the antenna gain by 3 dB, and partially shielding objects on the other side. If the antenna is too close to the conductive surface, the image currents cancel the currents in the antenna, resulting in poor radiation efficiency. This problem is often addressed by including a quarter-wavelength space between the radiating element and the ground plane, but such a structure then requires a minimum thickness of λ/4.Since the antennas are considered to work in the UHF-range, the minimum thickness would be around 8 cm, which is not reasonable.

Metamaterial AMC Reflectivitiy “+1“ (reflection magnitude 1 and reflection phase 0°) Can be achieved by utilizing periodic patch with via geometry or by planar achitecture without the need of vias  FSS (Frequency Selective Surface) GA (Genetic algorithm) for an optimized FSS unit cell size,geometry and dielectric constant and thickness of the substrate material by a planar architecture, without the need for vias, which incorporates a high impedance Frequency Selective Surface (FSS) into its design. The EBG surface is obtained by utilizing a Frequency Selective Surface (FSS) on top of a thin dielectric substrate backed by a metallic ground plane, to act as an Artificial Magnetic Conductor (AMC). The antenna is then placed above the EBG surface to create the integrated EBG conformal antenna (Fig.).

Metamaterial - Simulations FSS-design GA-Output Parameter For fabrication purposes, the designs were limited to a maximum dielectric thickness of 5.0 mm, a maximum dielectric constant of 12.2, and a maximum unit cell size of 5.6 cm what correspond to ~ λ/5 (at 910 MHz). The unit cell is always a square and contains 8-fold symmetry in order to guarantee the same response to a TE or a TM polarized incident wave. The above parameters will allow for a relatively thin, inexpensive structure to be fabricated that has a small unit cell size with respect to wavelength. We tried in the following simulations to decrease the unit cell size as well as the maximum dielectric thickness on condition to get an AMC ground plane. Another constraint was the total size of the ground plane; normally it should be a few wavelengths but as the antenna should be as small as possible to getting attach to even small objects the size of the ground has to be limited to ~10-11 cm (correspond to ~ λ/3 at the desired frequency range (860 MHz-960 MHz)). To build the AMC-ground plane an array of 2 x 2 unit cells has been used (Fig.). Due to this size limitation, the surface wave won’t be prohibited and the directivity won’t be increased in this case. Unit Cell FSS-Screen with Antenna Antenna Structure

Metamaterial – Simulations (cont‘d) Input Impedance- Comparison (with and without using an FSS-layer) Resonance frequency is decreasing by using a FSS-layer

Metamaterial – Simulations (cont‘d) Bandwidth and maximum Directivity - Comparison (with and without using an FSS-layer) Higher Power transfer for the case with the FSS-layer (~94 % instead of 83 %) Directivity is alternating in the range between 900 and 950 MHz around 1.2 Antenna could be made smaller  future work Bandwidth sligthly smaller for the FSS-case As before, the FSS-layer was removed to examine its influence on the antenna-system. The resonance frequency was shifted again, but within the considered range (860 -960 MHz), the input impedance is pretty similar. That is why it might have been possible to shrink the size of the antenna by a factor of 1.05. Determining the extent to which the antenna size will change is an area for future work (see next chapter “Summary and Future work”).The directivity again shows some alteration (around 1.2) in the range between 900 and 950 MHz. Two peaks appear on the power plot in the range between 910 MHz and 930 MHz. By comparing the input impedances, we observe that there is a kind of resonance frequency that could be due to the FSS-layer that was originally supposed to resonate in this range. A higher peak is achieved by using an FSS-layer, but the bandwidth is slightly smaller in this case. A reason for this could be the very narrow frequency range, where we have AMC.  

Metamaterials (cont‘d) Summary Using an FSS-layer drop the resonance frequency Changing the size of the antenna to get a desired input impedance is very difficult Directivity behaviour changes sligthly Higher power delivered to the antenna with the FSS-layer Bandwidth slighly smaller for the FSS-case Future work Increasing Bandwidth Changing structure of the AMC instead of the antenna size Antenna attached to metal objects  Performance will change Tunable antenna design  provide tolerance for fabrication

Fabrication of AMCs GA Input Parameter Output Configuration of AMC

FSS Layer FSS Unit Cell /2 x  /2 FSS Layer Reflection Crosses 0 at 939 MHz

Directivity AMC Dual-Band PIFA Inductively Coupled PIFA Directivity   Directivity 867 MHz 940 MHz PEC Ground 2.5255 2.6126 AMC Ground 2.892 3.125 Increased 0.3665 0.5124   Directivity 915 MHz 940 MHz PEC Ground 2.7074 2.421 AMC Ground 3.6855 2.9899 Increased 0.9781 0.5689

Radiation [867 MHz] AMC Dual-Band PIFA Inductively Coupled PIFA PEC

Optimization Dual-Band PIFA Design

Optimization Dual-Band PIFA Design STUB 2.8 Imaginary 867 MHz [W] Real Power 867 MHz [%] Gap 2.6 157.45 40.29 63.56 Gap 2.5 142.94 3.55 29.90 Gap 2.4 153.95 26.54 77.39 Gap 2.2 131.14 2.66 10.70 Gap 2.1 128.29 2.69 9.23 Note: 915 MHz and 940 MHz were not able to be sufficiently matched.

Optimization Inductively-Coupled Feed Loop PIFA Design

Optimization Inductively-Coupled Feed Loop PIFA Design Dimension [mm] Imaginary 915 MHz [W] Real 915 MHz [W] Imaginary 940 MHz [W] 940 MHz [W] Power 915 MHz [%] 940 MHz [%] 8.90 164.15 0.07 169.24 3.27 0.92 17.11 9.10 164.23 0.06 170.52 4.52 0.78 20.97 9.20 162.70 0.24 166.82 1.47 3.64 9.68 9.25 166.60 2.98 171.89 0.51 26.83 2.44 9.30 158.90 0.14 163.86 1.29 3.11 10.70 9.40 163.04 0.13 167.28 4.80 1.90 9.50 163.52 0.17 168.06 4.42 2.34 23.91

OBSERVATIONS Dual-Band PIFA Design showed to be platform tolerant in numerous cases Inductively Coupled Feed Loop PIFA was very sensitive to platform An optimization was done for each mounting material with the Inductively Coupled Feed Loop PIFA The AMC ground plane did significantly improve the directivity and reduce the backlobe in both antenna cases An optimization needed to be done using the AMC for both antenna cases because the impedance was altered The Dual-Band PIFA Design was optimized to sufficient operation but the Inductively Coupled Feed Loop PIFA was not

ALTERNATE PLATFORM-TOLERANT RFID DESIGNS *courtesy of Prof. K.W.Leung City University of Hong Kong

- Background Information RFID Tag Design - Background Information Inductive-coupled feeding design

Platform Tolerant Principle -Using a patch antenna as the resonating element -The Tag antenna and the surface material are isolated by the ground plane -The Tag has a stable performance regardless of the mounting surface

First Design Size of Ground Plane: 83.638 x 112.058mm (0.26λx 0.34λ) RFID Tag Antenna Configuration First Design Size of Ground Plane: 83.638 x 112.058mm (0.26λx 0.34λ) Size of Printed Antenna: 64.4 x 89.95mm(0.2λ x 0.27λ) Substrate thickness: 1.524mm Substrate dielectric constant: 3.38 Substrate loss tangent: 0.0021

Transmitted power of the Reader - 1W (30dBm) RFID Tag Antenna Configuration Chip Impedance - 20.83 – j116.67Ω Transmitted power of the Reader - 1W (30dBm) Gain of the Reader antenna - ~7.5dBi

Simulated Antenna Gain RFID Tag Antenna Configuration Simulated Antenna Gain Gain: ~ -8.6 to -0.77 dBi

Current Distribution (First Design) RFID Tag Antenna Configuration Current Distribution (First Design) 902 MHz

Current Distribution (First Design) RFID Tag Antenna Configuration Current Distribution (First Design) 915 MHz

Current Distribution (First Design) RFID Tag Antenna Configuration Current Distribution (First Design) 928 MHz

Result and Analysis Measurement Method -The Read Range was measured in the EMC Chamber -Reader Antenna was moved inside the EMC Chamber -Measure the maximum readable distance that the signal can be detected

RFID Tag Antenna Configuration Measurement Method - RFID Tag was fixed by the foam stand and measured at different orientation angles (0 deg, 45 deg, 90 deg) 0 deg 45 deg 90 deg

Tag Antenna Configuration (by inductively-coupled feeding) Result and Analysis Tag Antenna Configuration (by inductively-coupled feeding) -The tag has similar performances for different angles

Result and Analysis Second Design -A Philips’s chip SL3S10 01 FTT is used -Impedance of the chip: 16 – j380Ω (much more capacitive) -Difficult to match using the first design -Introduce a new feed network

Second Tag Antenna Design RFID Tag Antenna Configuration Second Tag Antenna Design - Directly connect the feed network to the radiating patch at several point Size of Ground Plane: 83.638 x 112.058mm (0.26λx 0.34λ) Size of Printed Antenna: 54.45 x 93.3mm(0.17λ x 0.28λ)

Current Distribution (Second Design) RFID Tag Antenna Configuration Current Distribution (Second Design) 902 MHz

Current Distribution (Second Design) RFID Tag Antenna Configuration Current Distribution (Second Design) 915 MHz

Current Distribution (Second Design) RFID Tag Antenna Configuration Current Distribution (Second Design) 928 MHz

Platform Tolerance Test Result and Analysis Platform Tolerance Test -Following surfaces were used in the test: -Acrylic (200 x 200 x 3 mm) -Wood (200 x 200 x 3 mm) -Aluminium (200 x 200 x 3 mm)

Result and Analysis Platform Tolerance -The tag has stable performance over different surfaces -The longest read range is obtained for the metal (Aluminium) case because EM wave is reflected by the metal