Game-Theoretic Network Bandwidth Distribution for Self- Adaptive - - PowerPoint PPT Presentation

▶
game theoretic network bandwidth distribution for self
SMART_READER_LITE
LIVE PREVIEW

Game-Theoretic Network Bandwidth Distribution for Self- Adaptive - - PowerPoint PPT Presentation

Dec 30, 2023 239 likes •591 views

Game-Theoretic Network Bandwidth Distribution for Self- Adaptive Cameras Gautham Nayak Seetanadi 1 Martina Maggio, Karl-Erik rzen 1 Luis Almeida 2 Luis Oliveira 3 1 Department of Automatic Control, Lund University 2 Universidade do Porto /

slide-1
SLIDE 1

Game-Theoretic Network Bandwidth Distribution for Self- Adaptive Cameras

Gautham Nayak Seetanadi1 Martina Maggio, Karl-Erik Årzen1 Luis Almeida2 Luis Oliveira3

1Department of Automatic Control, Lund University 2Universidade do Porto / Faculdade de Engenharia 3Department of Computer Science, University of Pittsburgh

slide-2
SLIDE 2

Introduction

  • Multiple adaptive camera network
  • Game-theoretic resource manager
  • Improved bandwidth utilisation
  • Local PI Control
  • Convergence guarantees on

bandwidth

slide-3
SLIDE 3

Motivation

  • Problem of Allocation
  • Finite Resource
  • Multiple Entities
  • Fair Bandwidth Distribution
  • Global Threshold Limit
  • Successfully applied to Multicore systems [1]

[1] M.Maggio et al, “A game-theoretic resource manager for rt applications”, In Euromicro Conference on Real-Time Systems, 2013

slide-4
SLIDE 4

Decentralised Resource Allocation

  • Resource Manager Allocates Resources
  • Cameras Choose Service Level
  • Change Quality

Data Control Camera 1

λ1

Camera 2 Camera n

λ2 λn

Resource Manager

Ethernet

slide-5
SLIDE 5

Decentralised Resource Allocation

  • Resource Manager Allocates Resources
  • Cameras Choose Service Level
  • Change Quality
  • Determines Responsibility

λi

Data Control Camera 1

λ1

Camera 2 Camera n

λ2 λn

Resource Manager

Ethernet

slide-6
SLIDE 6

Resource Manager

  • Game Theory Based
  • Matching function determines fairness

fp,w = Bp,w − sp,w Bp,w

slide-7
SLIDE 7

Resource Manager

slide-8
SLIDE 8

Resource Manager

Camera2 Camera1 Manager t = 0 t = 1 t = 2 50% 50%

slide-9
SLIDE 9

Resource Manager

Camera2 Camera1 Manager t = 0 t = 1 t = 2 50% 75% 50% 25%

slide-10
SLIDE 10

Camera

Original Frames Encoded Frames

slide-11
SLIDE 11

Camera

  • Camera generates an image of size
  • Adaptable quality
  • Normalised error

sp,w = 0.01 · qp,w · sp,max + δsp,w,

Can be

  • ve or +ve
slide-12
SLIDE 12

Camera Loop

Controller (kp, ki) Camera Quality (qp,w) −1

P

Allocated Bandwidth (Bp,w) Error (ep,w) Frame Size (sp,tw)

slide-13
SLIDE 13

Network Allocation

  • SDN using Flexible Time Triggered - SE [2]

[2] P .Pedreiras and L.Almeida, “The flexible time triggered paradigm. An approach to QOS management in distributed real-time systems”, 2003

Camera2 Camera1 Manager t = 0 t = 1 t = 2

slide-14
SLIDE 14

NETWORK ALLOCATION

  • SDN using Flexible Time Triggered - SE [2]

[2] P .Pedreiras and L.Almeida, “The flexible time triggered paradigm. An approach to QOS management in distributed real-time systems”, 2003

Camera2 Camera1 Manager t = 0 t = 1 t = 2 I1,1 I1,2 I2,1 I2,2

slide-15
SLIDE 15

NETWORK ALLOCATION

  • SDN using Flexible Time Triggered - SE [2]

[2] P .Pedreiras and L.Almeida, “The flexible time triggered paradigm. An approach to QOS management in distributed real-time systems”, 2003

Camera2 Camera1 Manager t = 0 t = 1 t = 2 I1,1 I1,2 I2,1 I2,2 I1,2 : q1,2 → s1,2 B1,t=0 → B1,w=2

slide-16
SLIDE 16

NETWORK ALLOCATION

  • SDN using Flexible Time Triggered - SE [2]

[2] P .Pedreiras and L.Almeida, “The flexible time triggered paradigm. An approach to QOS management in distributed real-time systems”, 2003

Camera2 Camera1 Manager t = 0 t = 1 t = 2 I1,1 I1,2 I1,4 I1,5 I2,1 I2,2 I2,3 I2,4 I1,2 : q1,2 → s1,2 B1,t=0 → B1,w=2

slide-17
SLIDE 17

Implementation

  • Cameras
  • Logitech c270. COTS
  • OpenCV
  • Network
  • Deadlines enforced using Flexible Time Triggered

Switched Ethernet (FTT-SE)

  • i7-4790 8 core PC running Fedora
slide-18
SLIDE 18

Experimental Setup

Blue Camera Red Camera

slide-19
SLIDE 19

Assessment

  • Criteria?
  • Fair Bandwidth Usage (Manager)
  • Complete Bandwidth Utilisation (Camera)
  • SSIM (Structural Similarity Index)
  • SSIM
slide-20
SLIDE 20

Experiments

Resource Allocator Camera

Equal Bandwidth Distribution No Adaptation Equal Bandwidth Distribution DARTES Model [2] Equal Bandwidth Distribution PI Controller Game-Theoretic RA PI Controller

[2] J. Silvestre-Blanes, L. Almeida, R. Marau, and P . Pedreiras. “Online qos management for multimedia real-time transmission in industrial networks.” IEEE Transactions on Industrial Electronics, 58(3), March 2011

slide-21
SLIDE 21

Experiment 1

Resource Allocator Camera

Equal Bandwidth Distribution No Adaptation 10 20 30 40 50 60 70 80 0.0 10.0 20.0 30.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2

slide-22
SLIDE 22

Experiment 1

Resource Allocator Camera

Equal Bandwidth Distribution No Adaptation 10 20 30 40 50 60 70 80 0.0 10.0 20.0 30.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2

slide-23
SLIDE 23

10 20 30 40 50 60 70 80 0.0 10.0 20.0 30.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2

Experiment 1

Resource Allocator Camera

Equal Bandwidth Distribution No Adaptation

slide-24
SLIDE 24

Experiment 2

10 20 30 40 50 60 70 80 90 100 110 120 0.0 1.0 2.0 3.0 4.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2

Resource Allocator Camera

Equal Bandwidth Distribution DARTES Model

slide-25
SLIDE 25

Experiment 2

10 20 30 40 50 60 70 80 90 100 110 120 0.0 1.0 2.0 3.0 4.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2

Resource Allocator Camera

Equal Bandwidth Distribution DARTES Model

slide-26
SLIDE 26

Experiment 3

10 20 30 40 50 60 70 80 90 100 110 120 0.0 1.0 2.0 3.0 4.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2

Resource Allocator Camera

Equal Bandwidth Distribution PI Controller

slide-27
SLIDE 27

Experiment 3

10 20 30 40 50 60 70 80 90 100 110 120 0.0 1.0 2.0 3.0 4.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2

Resource Allocator Camera

Equal Bandwidth Distribution PI Controller

slide-28
SLIDE 28

Experiment 4

5 10 15 20 25 30 35 40 45 50 55 60 0.0 1.0 2.0 3.0 4.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2

Resource Allocator Camera

Game-Theoretic RA PI Controller

slide-29
SLIDE 29

Experiment 4

5 10 15 20 25 30 35 40 45 50 55 60 0.0 1.0 2.0 3.0 4.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2

Resource Allocator Camera

Game-Theoretic RA PI Controller

slide-30
SLIDE 30

SSIM

200 400 600 800 1000 1200 1400 1600 0,05 0,1 Frame Number SSIM with PI controller − SSIM with DARTES model

Camera 1 Camera 2

slide-31
SLIDE 31

SSIM

200 400 600 800 1000 1200 1400 1600 0,05 0,1 Frame Number SSIM with PI controller − SSIM with DARTES model

Camera 1 Camera 2

slide-32
SLIDE 32

Conclusions

  • CPU allocation strategy
  • Need for adaptation
  • PI control advantages
  • Game-Theoretic resource manager efficiency
  • Convergence guarantees
slide-33
SLIDE 33

Future Work

  • Time triggered network manager decisions have

drawbacks

  • Moving to event triggered
slide-34
SLIDE 34

gautham@control.lth.se

10 20 30 40 50 60 70 80 90 100 110 120 0.0 1.0 2.0 3.0 4.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2 10 20 30 40 50 60 70 80 90 100 110 120 0.0 1.0 2.0 3.0 4.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2 5 10 15 20 25 30 35 40 45 50 55 60 0.0 1.0 2.0 3.0 4.0 Time [s] BW [Mbps] AllocBW Camera 1 AllocBW Camera 2 InstBW Camera 1 InstBW Camera 2

Resource Allocator Camera