An Adaptive Tree Algorithm to Approach Collision-Free Transmission - - PowerPoint PPT Presentation

▶
an adaptive tree algorithm to approach collision free
SMART_READER_LITE
LIVE PREVIEW

An Adaptive Tree Algorithm to Approach Collision-Free Transmission - - PowerPoint PPT Presentation

Nov 13, 2023 473 likes •799 views

An Adaptive Tree Algorithm to Approach Collision-Free Transmission in Slotted ALOHA Molly Zhang, Luca de Alfaro, JJ Garcia-Luna Aceves University of California, Santa Cruz Outline Problem Statement Adaptive Tree ALOHA Performance Setting:

slide-1
SLIDE 1

An Adaptive Tree Algorithm to Approach Collision-Free Transmission in Slotted ALOHA

Molly Zhang, Luca de Alfaro, JJ Garcia-Luna Aceves University of California, Santa Cruz

slide-2
SLIDE 2

Outline

Problem Statement Adaptive Tree ALOHA Performance

slide-3
SLIDE 3

Setting: Time-Slotted Channel Access

User 1 User 2 User 3 Channel time

slide-4
SLIDE 4

Goal: Learning Coordination in Channel Access

  • Turn Taking
  • High Network Utilization
  • Avoid collisions
  • Avoid empty time slots
slide-5
SLIDE 5

History: ALOHA

User 1 User 2 User 3 Channel time

ALOHA protocol: Transmit when you like, and if there are collisions, retry. Max utilization ≈ 18%

slide-6
SLIDE 6

History: Slotted ALOHA

Slotted ALOHA protocol: Time divided to time

  • slots. Transmit at the

beginning of time slots. Max utilization ≈ 37%

User 1 User 2 User 3 Channel

slide-7
SLIDE 7

History: Slotted ALOHA with Exponential Backoff

Exponential Backoff

  • Transmit with probability p
  • Collision: halves p
  • Success: doubles p

Max utilization ≈ 100% (very unfair condition)

slide-8
SLIDE 8

Goal: Learning Coordination in Channel Access

1 2 3

Can we do better? Can nodes learn to coordinate with Reinforcement Learning or Machine Learning?

slide-9
SLIDE 9

Reinforcement Learning and Expert-based Learning

slide-10
SLIDE 10

ALOHA-Q: Choosing transmission slot [Chu et al, 2012]

  • Learn the weight of slots in a frame.
slide-11
SLIDE 11
  • Learn the weight of slots in a frame.
  • Transmit in the highest-weight slot

ALOHA-Q: Choosing transmission slot [Chu et al, 2012]

slide-12
SLIDE 12
  • Learn the weight of slots in a frame.
  • Transmit in the highest-weight slot
  • Different nodes learns different slot

ALOHA-Q: Choosing transmission slot [Chu et al, 2012]

Transmissions

slide-13
SLIDE 13

ALOHA-Q: Choosing transmission slot [Chu et al, 2012]

Problems:

  • Frame length N selection
  • Slow learning

Transmissions

slide-14
SLIDE 14

(3, 2) (1, 2) (2, 2) (0, 2) (0, 1) (0, 0) (1, 1)

AT-ALOHA

Guide learning and conflict resolution via a policy tree.

(i, m) : transmit at time i every 2m slots

slide-15
SLIDE 15

(3, 2) (1, 2) (2, 2) (0, 2) (0, 1) (0, 0) (1, 1)

AT-ALOHA

Guide learning and conflict resolution via a policy tree.

slide-16
SLIDE 16

(1, 1) (3, 2) (1, 2) (2, 2) (0, 2) (0, 1) (0, 0)

AT-ALOHA

Guide learning and conflict resolution via a policy tree.

slide-17
SLIDE 17

(1, 1) (3, 2) (1, 2) (2, 2) (0, 2) (0, 1) (0, 0)

AT-ALOHA

Guide learning and conflict resolution via a policy tree.

(1, 2) Every child transmits half the times of the parent.

slide-18
SLIDE 18

(1, 1) (3, 2) (1, 2) (2, 2) (0, 2) (0, 1) (0, 0)

AT-ALOHA

Guide learning and conflict resolution via a policy tree.

(1, 2)

  • Nodes that are not one the descendant of the other do

not conflict.

  • Conflicts are rare. Coordination is facilitated.
slide-19
SLIDE 19

AT-ALOHA

Different nodes learn a different tree to co-exist conflict-free

slide-20
SLIDE 20

Next: How do the AT-ALOHA nodes learn different trees ?

slide-21
SLIDE 21

AT-ALOHA Update: Demotion After Collision

p=0.5 p=0.5

slide-22
SLIDE 22

AT-ALOHA Update: Demotion After Collision

p=0.5 p=0.5

slide-23
SLIDE 23

AT-ALOHA Update: barge into empty slots

p 1 − p = nodes that could have transmitted in time slot The barge-in probability p is tuned based on the number of active nodes in a network.

slide-24
SLIDE 24

AT-ALOHA Update: Normalization

merge sibling nodes remove redundant descendants

slide-25
SLIDE 25

AT-ALOHA Update Pruning to max depth and max number of nodes

slide-26
SLIDE 26

AT-ALOHA: additional tuned parameters

➔ Maintaining 5% empty slots

◆ “Transmission Tax”: a node has to give up its transmission policy at a small probability

➔ Maintaining a constant (1.4) empty-to-collision ratio

◆ By tuning barge-in probability ◆ Maximize likelihood of only one transmitting into empty slot

slide-27
SLIDE 27

AT-ALOHA Performance Metric

  • Network Utilization: Ratio of successful transmission
  • Fairness Metric: Jain index
slide-28
SLIDE 28

AT-ALOHA Performance

  • 10 nodes -> 50 nodes -> 30 nodes
  • High Utilization and Low Empty slots
  • r Collisions throughout
slide-29
SLIDE 29

AT-ALOHA Performance comparison

  • AT-ALOHA
  • EB-ALOHA: ALOHA with exponential

backoff

  • EB-ALL-ALOHA: ALOHA with exponential

backoff applied to all nodes

  • ALOHA-Q: Chu et al.

AT-ALOHA has both high network utilization and high fairness under varying network conditions

Network Utilization Fairness

slide-30
SLIDE 30

Conclusions

  • We introduced a “Adaptive Tree” ALOHA protocol.
  • Learns to maintain high utilization and fairness under varying network

condition

(1, 1) (3, 2) (1, 2) (2, 2) (0, 2) (0, 1) (0, 0) (1, 2)

slide-31
SLIDE 31

Thank you!