A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the - - PowerPoint PPT Presentation

▶
a new dynamic frame slotted aloha anti collision
SMART_READER_LITE
LIVE PREVIEW

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the - - PowerPoint PPT Presentation

Oct 08, 2022 295 likes •696 views

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Rafael Perazzo Barbosa Mota IME - USP 05 de junho de 2013 Agenda 1 Motivation and relevance 2 Background 3 Related Work 4 Our Proposal 5 Results and

slide-1
SLIDE 1

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things

Rafael Perazzo Barbosa Mota

IME - USP

05 de junho de 2013

slide-2
SLIDE 2

Agenda

1 Motivation and relevance 2 Background 3 Related Work 4 Our Proposal 5 Results and discussion 6 Conclusions 7 References

slide-3
SLIDE 3

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Motivation and relevance

Table of Contents

1 Motivation and relevance 2 Background 3 Related Work 4 Our Proposal 5 Results and discussion 6 Conclusions 7 References

3 / 31

slide-4
SLIDE 4

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Motivation and relevance

Motivation and relevance

Radio Frequency Identification (RFID) is a key technology of IoT since small passive RFID tags make it possible to link millions and billions of physical products with Internet [1].

2 / 31

slide-5
SLIDE 5

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Motivation and relevance

Motivation and relevance

Therefore, RFID tag anticollision algorithms will play an important role in IoT [1].

3 / 31

slide-6
SLIDE 6

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Motivation and relevance

Background - DFSA

4 / 31

slide-7
SLIDE 7

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Background

Table of Contents

1 Motivation and relevance 2 Background 3 Related Work 4 Our Proposal 5 Results and discussion 6 Conclusions 7 References

5 / 31

slide-8
SLIDE 8

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Background

Dynamic Frame Slotted Aloha - DFSA

Algorithm 1 DFSA algorithm Require: L ⊲ L is the initial frame size

1: continue ← true 2: n ← L 3: repeat

⊲ While collisions occurs

4:

i ← 0 ⊲ Initial slot time

5:

counter ← 0 ⊲ Number of received replies (=1, =0 or >1)

6:

collisions ← 0 ⊲ Collision counter

7:

for i ≤ n do ⊲ Sends every slot time

8:

Query(n,i) ⊲ Sends a Query Command with frame size n and slot i

9:

Wait for reply

10:

if (counter == 1) then

11:

QueryRep() ⊲ Reader sends an ACK to identify the tag

12:

else if (counter > 1) then

13:

collisions ← colisions + 1

14:

end if

15:

end for

16:

if (collisions == 0) then

17:

continue ← false

18:

else

19:

n ←Call a function to calculate the next frame size

20:

L ← n

21:

end if

22: until (continue==true) 6 / 31

slide-9
SLIDE 9

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Related Work

Table of Contents

1 Motivation and relevance 2 Background 3 Related Work 4 Our Proposal 5 Results and discussion 6 Conclusions 7 References

7 / 31

slide-10
SLIDE 10

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Related Work

Q Algorithm [2]

8 / 31

slide-11
SLIDE 11

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Related Work

Schoute [3]

Algorithm 2 Schoute algorithm

1: function Schoute(collisions ) 2:

return round(2.39 ∗ collisions)

3: end function

9 / 31

slide-12
SLIDE 12

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Related Work

Mota 2013

Algorithm 3 Mota algorithm

1: function mota(collisions ) 2:

return round(2.62 ∗ collisions)

3: end function

10 / 31

slide-13
SLIDE 13

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Related Work

Eom-Leee [4]

Algorithm 4 Eom-Lee Estimation Algorithm

1: function estimation_eomlee( ǫ, collisions, success )

⊲ collisions and success are the number of collision and success slots in last frame, respectively. ǫ is the stop criteria

2:

b1 ← ∞

3:

y1 ← 2

4:

backlog ← L

5:

repeat

6:

bprox ← backlog y1 ∗ collisions + success

7: 8:

yprox ← 1 − e

−1 bprox

bprox ∗ (1 − (1 +

1 bprox ) ∗ e

−1 bprox )

9: 10:

backlog ← yprox ∗ collisions

11:

temp ← y1

12:

y1 ← yprox

13:

b1 ← bprox

14:

until (|y1 − temp| < ǫ)

15: return round(backlog) 16: end function

11 / 31

slide-14
SLIDE 14

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Related Work

Others

Dynamic, Adaptative and Splitting BTSA (Excellent System

  • Efficiency. Many changes must be done on tags operation) [1]

12 / 31

slide-15
SLIDE 15

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Related Work

Others

Dynamic, Adaptative and Splitting BTSA (Excellent System

  • Efficiency. Many changes must be done on tags operation) [1]

Vogt (equivalent to Q Algorithm) [5]

12 / 31

slide-16
SLIDE 16

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Related Work

Others

Dynamic, Adaptative and Splitting BTSA (Excellent System

  • Efficiency. Many changes must be done on tags operation) [1]

Vogt (equivalent to Q Algorithm) [5] Chen (worse than Q Algorithm) [6]

12 / 31

slide-17
SLIDE 17

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

Table of Contents

1 Motivation and relevance 2 Background 3 Related Work 4 Our Proposal 5 Results and discussion 6 Conclusions 7 References

13 / 31

slide-18
SLIDE 18

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

EDFSA-I

Estimate initial frame size L slot=0 START Send Query command and current slot Tags replies? 1 >1 idle++ slot++ collisions++ slot++ End of Frame ? End of Frame ? Send ACK slot++ Calculate new Frame Size using Eom-Lee method slot=0 Yes Yes No No

14 / 31

slide-19
SLIDE 19

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

EDFSA-II

15 / 31

slide-20
SLIDE 20

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

EDFSA-II

Algorithm 5 Estimated Dynamic Framed Slotted Aloha - EDFSA Require: ǫ ⊲ ǫ is the stop criteria

1: L ← estimation(1, 3) ∗ 0.67 2: i ← 1

⊲ Initial slot time

3: counter ← 0

⊲ Number of received replies (=1, =0 or >1)

4: collisions ← 0

⊲ Collision counter

5: for (i = 1; i ≤ L; i ← i + 1) do

⊲ Sends every slot time

6:

Query(L,i) ⊲ Sends a Query Command with frame size n and slot i

7:

Wait for reply

8:

if (counter == 1) then

9:

QueryRep() ⊲ Reader sends an ACK to identify the tag

10:

success ← success + 1

11:

else if (counter > 1) then

12:

collisions ← collisions + 1

13:

resolve_collisions() ⊲ Collisions are resolved as soon as they

  • ccur

14:

end if

15: end for

16 / 31

slide-21
SLIDE 21

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

How to estimate the initial frame [7] ?

Send i times QueryEst command

collisions=i or idle=i collisions=0 idle=0 success=0

START Q++

  • r

Q-- Yes No FinalQ=Q Send i times QueryEst

collisions=i or idle=i

Yes No FinalQ is the estimated number of tags END

17 / 31

slide-22
SLIDE 22

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

How to resolve local collisions ?

START L=3 Next slot Send Query with slot i Wait for replies Number of replies ? QueryRep() Next slot success++ collisions++ >1 =1 End of frame ? No Yes collisions = 0 ? END Yes L = mota(collisions)

18 / 31

slide-23
SLIDE 23

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

How to select best parameters for estimation ?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 200 400 600 800 1000 1200 1400 1600 1800 Best Initial Q Value Number of tags c=1 and i=3 c=1 and i=5 c=0.3 and i=3 c=0.2 and i=3 c=0.2 and i=5 19 / 31

slide-24
SLIDE 24

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

How to select best parameters for estimation ?

25 50 75 100 125 150 175 200 225 250 275 300 325 350 375 400 50 250 450 650 850 1050 1250 1450 1650 1850 Delay (slots) Number of tags c=1 and i=3 c=1 and i=5 c=0.3 and i=3 c=0.2 and i=3 c=0.2 and i=5 20 / 31

slide-25
SLIDE 25

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

How to select the best initial frame size for resolve local collisions ?

0 % 5 % 10 % 15 % 20 % 25 % 30 % 35 % 40 % 45 % 50 % 55 % 60 % 65 % 70 % 75 % 80 % 85 % 90 % 95 % 100 % 2 3 4 5 6 7 8 9 10 11 12 Frequency Number of tags in collision

Mean: aprox.(2.616) 2.62 (2.633) Sample size: 88751 collisions slots Confidence Interval (CI): 99% 60.2082 25.0518 9.5896 3.3430 1.1797 0.4124 0.1532 0.0338 0.0169 0.0090 0.0023 21 / 31

slide-26
SLIDE 26

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

How to select the next frame estimation method ?

−20 % 0 % 20 % 40 % 60 % 80 % 100 % 120 % 140 % 160 % 180 % 200 % 220 % 240 % 260 % 280 % 300 % 320 % 340 % 360 % 380 % 400 % 420 % 440 % 460 % 2 3 4 5 6 7 8 Average Difference compared to Q Algorithm Number of tags in collision

  • PS. Negative number means a decrease on System Efficiency
  • PS. Positive number means a increase on System Efficiency

Initial Frame Size: 3

Lower Bound Schoute Eom−Lee Mota

336 211 117 84 39 33 27 429 200 113 72 39 27 −3 371 195 122 72 57 37 23 357 242 113 80 46 30 27

22 / 31

slide-27
SLIDE 27

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Our Proposal

How to select the next frame estimation method ?

0.06 0.09 0.12 0.15 0.18 0.21 0.24 0.27 0.30 0.33 0.36 0.39 0.42 0.45 0.48 0.51 0.54 0.57 0.60 0.63 0.66 0.69 0.72 0.75 0.78 0.81 0.84 2 3 4 5 6 7 8 System Efficiency Number of tags in collision

Confidence Interval (CI) 95% Initial Frame Size: 3

Q Algorithm Lower Bound Schoute Eom−Lee Mota

0.66 0.56 0.51 0.43 0.44 0.41 0.37 0.64 0.65 0.49 0.45 0.41 0.39 0.38 0.61 0.59 0.50 0.46 0.39 0.40 0.38 0.74 0.57 0.49 0.43 0.39 0.38 0.29 0.14 0.19 0.23 0.25 0.28 0.30 0.30

23 / 31

slide-28
SLIDE 28

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Results and discussion

Table of Contents

1 Motivation and relevance 2 Background 3 Related Work 4 Our Proposal 5 Results and discussion 6 Conclusions 7 References

24 / 31

slide-29
SLIDE 29

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Results and discussion

System Efficiency

0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 System efficiency Number of tags Confidence Interval (CI) 95% Q Algorithm Schoute 128 Eom−Lee 128 EDFSA−I EDFSA−II 25 / 31

slide-30
SLIDE 30

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Results and discussion

Identification time (in timeslots)

500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 Mean identification time (slots) Number of tags Confidence Interval (CI) 95% Q Algorithm Schoute 128 Eom−Lee 128 EDFSA−I EDFSA−II 26 / 31

slide-31
SLIDE 31

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Results and discussion

Comparison

−30 % −28 % −26 % −24 % −22 % −20 % −18 % −16 % −14 % −12 % −10 % −8 % −6 % −4 % −2 % 0 % 2 % 4 % 6 % 8 % 10 % 12 % 14 % 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 Average Difference in number of slots(%) compared to Q Algorithm Number of tags Schoute 128 Eom−Lee 128 EDFSA−I EDFSA−II

−14.95 −13.40 −6.09 −3.76 −3.87 −3.45 −1.41 −1.12 −1.74 −1.12 0.13 −0.48 0.19 0.44 0.09 0.33 −0.41 0.17 −13.98 −12.44 −7.17 −6.82 −7.00 −8.44 −7.88 −0.64 −1.89 −1.76 −4.55 −6.19 −6.86 −7.07 −5.90 −6.84 −7.16 −6.25 5.74 −1.06 −2.98 −4.40 −7.96 −9.13 −9.38 −8.43 −8.82 −7.40 −8.58 −8.97 −9.65 −9.41 −9.38 −10.61 −11.54 −8.68 −19.40 −17.66 −18.38 −19.75 −22.34 −22.33 −22.36 −22.83 −23.02 −21.27 −21.90 −25.31 −24.32 −26.45 −25.11 −25.73 −15.71 −18.90

27 / 31

slide-32
SLIDE 32

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Conclusions

Table of Contents

1 Motivation and relevance 2 Background 3 Related Work 4 Our Proposal 5 Results and discussion 6 Conclusions 7 References

28 / 31

slide-33
SLIDE 33

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Conclusions

Conclusions

We presented two new anti-collision algorithms based on existing estimation methods;

29 / 31

slide-34
SLIDE 34

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Conclusions

Conclusions

We presented two new anti-collision algorithms based on existing estimation methods; The EDFSA-I has a gain up to 12% related to Q Algorithm;

29 / 31

slide-35
SLIDE 35

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Conclusions

Conclusions

We presented two new anti-collision algorithms based on existing estimation methods; The EDFSA-I has a gain up to 12% related to Q Algorithm; The EDFSA-II has a gain up to 26% related to Q Algorithm while Splitting BTSA has a gain up to 23% and Adaptative, Dynamic has a gain up to 14%;

29 / 31

slide-36
SLIDE 36

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Conclusions

Conclusions

We presented two new anti-collision algorithms based on existing estimation methods; The EDFSA-I has a gain up to 12% related to Q Algorithm; The EDFSA-II has a gain up to 26% related to Q Algorithm while Splitting BTSA has a gain up to 23% and Adaptative, Dynamic has a gain up to 14%; Both proposed methods make minor changes to default tag

  • perations and do not require additional resources, except an
  • ptional memory;

29 / 31

slide-37
SLIDE 37

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things Conclusions

Conclusions

We presented two new anti-collision algorithms based on existing estimation methods; The EDFSA-I has a gain up to 12% related to Q Algorithm; The EDFSA-II has a gain up to 26% related to Q Algorithm while Splitting BTSA has a gain up to 23% and Adaptative, Dynamic has a gain up to 14%; Both proposed methods make minor changes to default tag

  • perations and do not require additional resources, except an
  • ptional memory;

An ns-2 module with all proposed algorithms was also developed and is available freely for students and researchers.

29 / 31

slide-38
SLIDE 38

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things References

Table of Contents

1 Motivation and relevance 2 Background 3 Related Work 4 Our Proposal 5 Results and discussion 6 Conclusions 7 References

30 / 31

slide-39
SLIDE 39

A New Dynamic Frame Slotted Aloha Anti-Collision Algorithm for the Internet of Things References

References

[1]

  • H. Wu, Y. Zeng, J. Feng e Y. Gu, “Binary tree slotted aloha for passive rfid tag anticollision”,

Parallel and Distributed Systems, IEEE Transactions on, vol. 24, no 1, pp. 19–31, 2013, issn: 1045-9219. doi: 10.1109/TPDS.2012.120. [2]

  • V. Namboodiri, M. DeSilva, K. Deegala e S. Ramamoorthy, “An extensive study of slotted

aloha-based {rfid} anti-collision protocols”, Computer Communications, vol. 35, no 16, pp. 1955 –1966, 2012, issn: 0140-3664. doi: 10.1016/j.comcom.2012.05.015. endereço: http://www.sciencedirect.com/science/article/pii/S0140366412001776. [3]

  • F. Schoute, “Dynamic frame length aloha”, Communications, IEEE Transactions on, vol. 31, no 4,
  • pp. 565–568, 1983, issn: 0090-6778. doi: 10.1109/TCOM.1983.1095854.

[4] J.-B. Eom e T.-J. Lee, “Accurate tag estimation for dynamic framed-slotted aloha in rfid systems”, Communications Letters, IEEE, vol. 14, no 1, pp. 60–62, 2010, issn: 1089-7798. doi: 10.1109/LCOMM.2010.01.091378. [5]

  • H. Vogt, “Efficient object identification with passive rfid tags”, em Proceedings of the First

International Conference on Pervasive Computing, sér. Pervasive ’02, London, UK, UK: Springer-Verlag, 2002, pp. 98–113, isbn: 3-540-44060-7. endereço: http://dl.acm.org/citation.cfm?id=646867.706691. [6]

  • T. Li, S. Wu, S. Chen e M. Yang, “Energy efficient algorithms for the rfid estimation problem”, em

INFOCOM, 2010 Proceedings IEEE, 2010, pp. 1–9. doi: 10.1109/INFCOM.2010.5461947. [7]

  • Y. Cui e Y. Zhao, “A modified q-parameter anti-collision scheme for rfid systems”, em Ultra Modern

Telecommunications Workshops, 2009. ICUMT ’09. International Conference on, 2009, pp. 1–4. doi: 10.1109/ICUMT.2009.5345419.

31 / 31