A Discrete Particle Swarm Optimization for IoT services placement - - PowerPoint PPT Presentation

A Discrete Particle Swarm Optimization for IoT services placement over Fog infrastructures

Directed by:

• Patricia STOLF
• Thierry MONTEIL
• Jean-Marc PIERSON

PhD Student:

• Tanissia

DJEMAI

1

Outline

Experimental approach & results

• Methodology
• Results

Introduction

• Smart cities
• Internet of Things

applications

• Large scale

computing infrastructures

Problem formulation

• Fog hierarchical

Infrastructures

• IoT applications

graphs

• Objective

function

Strategies

• CloudOnly
• FogOnly
• FogCloud
• IoTCloud
• DCT
• DPSO

Conclusion

• Current Work
• Future

prospects

2

Introduction

(thesis context )

Strategies

Introduction

Problem formulation Experimental approach & results Conclusion 3

Heterogeneity Dynamicity Users number Energy greedy 25% 28% 47%

“Fog computing is a horizontal, physical or virtual resource paradigm that resides between smart end-devices and traditional cloud or data centers.” [NIST 2017]

Fog Infrastructures

Strategies Introduction

Problem formulation

Experimental approach & results Conclusion 4

IoT applications

Strategies Introduction

Problem formulation

Experimental approach & results Conclusion 5

S 2 S S 4 S 1 S 3 S S 2 S 1

Strategies Introduction

Problem formulation

Experimental approach & results Conclusion 6

S 2 S S 4 S 1 S 3 S S 2 S 1 S S 2 S 1 S S 1 S 2 S S 1 S 2 S 4

Energy & delay violation

Discrete Particle Swarm Optimization approach

Experimental approach & results Introduction Problem formulation

Strategies

Conclusion 7

Experimental approach & results Introduction Problem formulation

Strategies

Conclusion 8

START

1. Initialize all particles uniformly 2. Initialize velocities to 1 3. Evaluate fitness for each particle Xk 4. Update Personal best (pb) 5. Update ring neighbor best (nb) 6. Update velocity 7. Update particle position Is max iterations reached?

STOP

Yes No

Experimental approach & results Introduction Problem formulation

Strategies

Conclusion 9

IoTFogOnly CloudOnly FogCloud(FC) IoTCloud(IC) Discret Particle Swarm Optimization (DPSO) Dicothomous (DCT)

Experimental approach & results Introduction Problem formulation

Strategies

Conclusion 10

S S 2 S 1 S S 1 S 2

(1) Real Time (RT) (2) Mission Critical (MC) (3) Streamin (ST) (4) Best Effort (BE)

Experimental approach & results Introduction Problem formulation

Strategies

Conclusion 11

12

Conclusion (1)

1. Evolutionary approach and basic placement strategies. 2. DPSO gives a good tradeoff between energy and delay values. 3. Execution time. 4. Centralized approach. 5. Hierarchical topology. 6. Linear energy consumption profile. 7. Static infrastructure and VMs.

13

Experimental approach & results Introduction Problem formulation

Strategies

Conclusion

Conclusion (2)

14

Experimental approach & results Introduction Problem formulation

Strategies

Conclusion 15

Conclusion (3)

Solution quality impacted by time. Users mobility estimation Efficient handover and migrations approches Evaluation Services availability.

BIBLIOGRAPHY

[1] Z. A. Bonomi, Milito. Fog computing and its role in the internet of things.MCC’12, August 17, 2012, Helsinki, Finland, -1, 2012. [2] C. company. Cisco fog computing with iox.IEA 4E EDNA, Technology and Energy Assessment Report, -1, 2014. [3] L. L. Giang, Blackstock. Developing iot applications in the fog: a distributed dataflow approach.5th International Conference on the Internet of Things (IoT), 2015. [4] G. B. Gupta, Dastjerdi. ifogsim: A toolkit for modeling and simulation of resource management techniques in the internet

• f things, edge and fog computing environments.IEEE, -1, 2016. to appear.

[5] B. M. G. M. Iorga, Feldman. Fog computing conceptual model recommendations of the national institute of standards and technology.NIST Special Publication 500-325, -1, 2017.

16