Multi-Band Radio Frequency Energy Harvesting Storing in Super-Capacitor for Self- Sustainable Cognitive radio networks

Transcription

1 CREaTION Workshop Multi-Band Radio Frequency Energy Harvesting Storing in Super-Capacitor for Self- Sustainable Cognitive radio networks Luís M. Borges Fernando J. Velez 2005, it - instituto de telecomunicações. Todos os direitos reservados.

2 Outline Objectives and Motivation Specifications for the Radio Frequency Energy Harvesting and Super-Capacitor Storing system RF Energy Harvester Front End Block Design of N-stage Dickson Voltage Multiplier for DTT band Design of Super-Capacitor Storing System Energy Management Algorithm Experimental Results 5 and 7-stages RF Energy Harvester Radio Frequency Energy Harvesting and Storing system Conclusions Future work 2

3 Objectives and Motivation RF Energy Sources RF Energy Harvesting + Super-Capacitor Storing System Propose a novel dynamic energy management algorithm Propose a super-capacitor storing system that copes with low output voltages from the RF energy harvesting prototypes Propose RF energy harvesting Dickson voltage multiplier prototypes for different frequency bands such as Digital Terrestrial Television (DTT) The motivation is to create a self-sustainable system that harvests energy from electromagnetic waves and power-supply a cognitive radio node to send data opportunistically when the super-capacitor attains a certain level of voltage. 3

4 Specifications for the Radio Frequency Energy Harvesting and Super-Capacitor Storing system Depending on the type of embedded system which is going to be powered-up different requirements need to be fulfilled; Typical WSN node: V supply (min)=1.8 V; V supply (max)=3 V Cognitive Radio (CR) Node: * Overall power consumption of 200 mw and V supply =2-3 V are considered * M. Dardaillon, K. Marquet, T. Risset, and A. Scherrer, Software defined radio architecture survey for cognitive testbeds, in Proc. of the 8th International Wireless Communications and Mobile Computing Conference (IWCMC 2012), Limassol, Cyprus, August 2012, pp

5 Block diagram of the Radio Frequency Energy Harvesting and Super-Capacitor Storing system 5

6 V DC (V) Design of N-stage Dickson Voltage Multiplier for DTT band In a previous work it has been developed and tested 5-stage Dickson voltage multiplier prototypes for the GSM bands (900/1800); Due to the potential arising from the wide/broad deployment of DTT in Portugal we decided to develop RF energy harvesting and antennas to operate at these frequency bands ( MHz) stages 35 4-stages 5-stages 30 6-stages 7-stages 8-stages 25 9-stages stages 3-stages (w/o stub) 60 4-stages (w/o stub) 15 5-stages (w/o stub) 50 6-stages (w/o stub) 10 7-stages (w/o stub) 8-stages (w/o stub) stages (w/o stub) 10-stages (w/o stub) P RF (dbm) 20 η 0 (%) 10 3-stages 4-stages 5-stages 6-stages 7-stages 8-stages 9-stages 10-stages 3-stages (w/o stub) 4-stages (w/o stub) 5-stages (w/o stub) 6-stages (w/o stub) 7-stages (w/o stub) 8-stages (w/o stub) 9-stages (w/o stub) 10-stages (w/o stub) P (dbm) RF 6

7 30 Design of N-stage Dickson Voltage Multiplier for DTT band 25 Output Voltage 20 V DC (V) stages (Open Circuit) 5 5-stages (Short Circuit) 7-stages (Open Circuit) 7-stages (Short Circuit) P (dbm) RF 40 Conversion Efficiency stages (Open Circuit) 5-stages (Short Circuit) 7-stages (Open Circuit) 7-stages (Short Circuit) η 0 (%) P RF (dbm)

8 Design of N-stage Dickson Voltage Multiplier for DTT band For the 5 and 7 stage Dickson voltage multiplier prototypes different Half-Wave Dipole antennas for DTT band were designed: S 11 = db 8

9 Design of Super-Capacitor Storing System Since we are dealing with low voltages outputs from the RF energy harvesting prototype we decided to implemented a Buck-Boost converter controlled by an Arduino; The key principle of the boost converter is the ability of the inductor to resist changes in current by creating and destroying a magnetic field. In a boost converter, the output voltage is always higher than the input voltage. 9

10 Design of Super-Capacitor Storing System In order to meet the requirements aforementioned for our project some calculations and tests were made to find the best value for the inductor, Pre-charge capacitor and the current limiting resistor of the inductor; Trial cases to evaluate the precharge capacitors and inductors: 10

11 Design of Super-Capacitor Storing System Buck-boost in action (trial case 6) 11

12 Choice of the best values for electronic components of the Super-Capacitor Storing System To choose the best values for the electronic components of the Storing system diferent metrics were evaluated: Average charging voltage in C super-cap ; Average charging voltage per cycle in C super-cap ; Normalized charging voltage in C super-cap ; Normalized Charging Time per Cycle of C super-cap ; Low Traffic Ambient experiment with 5-Stage RF GSM Energy Harvester; Time Estimation to attain different C super-cap Harvesting Device and worst case scenario). Voltages (w/ 12

13 Average charging voltage in C super-cap 13

14 Average charging voltage per cycle in C super-cap Consider 2000 cycles of charging and discharging cycles for the Pre-charge capacitor. 14

15 Normalized charging voltage in C super-cap Does not take into account that as the energy builds up in the super-capacitor it takes more time to add more energy to the supercapacitor. =2000 max max _ 15

16 Normalized Charging Time per Cycle of C super-cap Longer time but higher voltage in C super-cap 16

17 Low Traffic Ambient experiment with 5-Stage RF Energy Harvester Assess the average voltage that each RF energy harvester prototype is able to deliver when placed in different traffic ambient; First experiments considered the 5-stage RF energy Harvester that scavenges from the GSM (900 and 1800) frequency bands that charge the pre-charge capacitor employed in the proposed storing system; Worst case scenario: Low Traffic Ambient with low number of celular devices/users (at most two users); 17

18 Time Estimation to attain different C super-cap Voltages (w/ Harvesting Device and worst case scenario) Estimated charging time and number of cycles for trial cases 6 and 8: 18

19 Energy Management Algorithm Required Voltage Supply vs Clock Frequency Sleep mode power consumptions: SLEEP MODE IDLE: 15 ma; SLEEP MODE ADC: 6.5 ma; SLEEP MODE PWR SAVE: 1.62 ma; SLEEP MODE EXT STANDBY: 1.62 ma; SLEEP MODE STANDBY: 0.84 ma; SLEEP MODE PWR DOWN: 0.36 ma. 19

20 Experimental results for the 5 and 7-stages TDD RF Energy Harvesters Output voltage 20

21 Experimental results for the 5 and 7-stages TDD RF Energy Harvesters (cont.) Conversion Efficiency 21

22 Complete solution of Radio Frequency Energy Harvesting and Storing system The parameters for the configuration of Radio Frequency Energy Harvesting and Storing system are the following: C pre-cap =1000 µf C super-cap =5 F V s-cap-charge-max =5 V 7-stage (open stub) RF Energy Harvester prototype; The program is loaded into the Arduino platform to control the energy transference is the one presented except that it does not considers deep sleep mode yet; Different power levels have been injected (by means of a signal generator) into the 7-stage RF energy harvester, i.e., -10 dbm, -5 dbm and -2 dbm at 754 MHz. 22

23 Complete solution of Radio Frequency Energy Harvesting and Storing system Experimental Apparatus 7-stage RF Energy Harvester Arduino UNO Super-capacitor storing system Opto-Coupler Board 23

24 Complete solution of Radio Frequency Energy Harvesting and Storing system experimental results V super-cap_initial (P RF =-10 dbm)= 0.52 V V super-cap_initial (P RF =-5 dbm)= 0 V V super-cap_initial (P RF =-2 dbm)= 0.61V V super-cap_max = V V super-cap_max = V V=0.318 V V= V V super-cap_max = V V= V 24

25 Conclusions Four RF energy harvester prototypes were projected and built to scavenge RF energy within the DTT frequency band; A super-capacitor storing system with adaptive duty cycle and voltage rates adjustment was projected and built; Different electronic components values were tested to find the best values for the Buck-Boost converter; The combinations of component values C pre-charge = 470 µf; L=101 µh and C pre-charge = 1000 µh; L=101 uh for the trial case number 6 and 8 are the ones that are more suitable; In case the Low Ambient Traffic scenario is the most common scenario, it is better to charge the super-capacitor initially and use the harvested energy to maintain the voltage of the supercapacitor(s) bank for longer time while it supplies the cognitive radio node. 25

26 Future work Propose an opportunistic electromagnetic RF Energy harvesting MAC protocol for CR networks; Use optimization algorithms to find best parameter values for the switching times of buck-boost converter; Perform experiments with multiple RF Energy Harvester prototypes supplying the super-capacitor storing system; Perform real environment experiments with RF Energy Harvester jointly with the super-capacitor storing system; Test the IC LTC3108 with the N-stage RF energy harvester prototypes to charge the super-capacitor(s) bank. 26

27 Thank you, Questions are Welcome 27