1 of 29

Machine Learning Empowered LoRaWAN for Command and Control (C2) Agility

Jahangir H. Sarker* and Ahmed Ghanmi**

*Canadian Joint Warfare Centre **Defence Research and Development Canada Department of National Defence Ottawa, Ontario, Canada

PLACE L2 HERE (DG LEVEL)

2 of 29

Outline of the Presentation

  • Introduction
  • System Model
  • Performances of Conventional Multichannel Slotted ALOHA System
  • Performance Analysis of Pure and Slotted ALOHA
    • 1. Multichannel pure ALOHA System
    • 2. Multichannel Slotted ALOHA System
  • Improved Normalized Readiness with machine learning empowered fast retransmission trials and the Slotted ALOHA (MLeFRSA)
  • Discussion
  • Conclusions and future works

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

2/(29)

3 of 29

Introduction 1/(4)

No single C2 approach is applicable for all missions and/or circumstances

more complex missions or circumstances require more network-enabled C2

27th ICCRTS October 25-27, 2022

Allocation of Decision Rights

Patterns of Interaction

Endeavour Complexity

Appropriate C2 Approach

Low

De-conflicted

Medium

Coordinated

Medium-High

Collaborated

High-Very high

Edge

3/(29)

Distribution of Information

4 of 29

Introduction 2/(4)

No single C2 approach is applicable for all missions and/or circumstances

One should seek an integrated design of the composite networks.

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

Problems?

1. Difficult to implement multiple-networks, especially for remote locations and emergency purposes

2. For choosing the appropriate integration of networks, mostly Throughput and Delay are considered.

Cost and Complexity will be very high.

4/(29)

5 of 29

Introduction 3/(4)

No single C2 approach is applicable for all missions and/or circumstances

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

Solution to reduce the Cost and Complexity

1. A single ad hoc network with multiple communications protocols having different efficiency is the solution.

Long range low power wide area network (LoRaWAN) wireless Network

Endeavour Complexity

Appropriate C2 Approach

Low

De-conflicted

Medium

Coordinated

Medium-High

Collaborated

High-Very high

Edge

P1

P2

P3

P4

5/(29)

6 of 29

Introduction 4/(4)

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

Throughput performance

Delay performance

Threat-hazard resiliency performance

6/(29)

Attacking drones

Swarm technology

Electronic warfare (EW) system adversary.

Communications and information systems (CIS)

7 of 29

System Model and Performances of Conventional Multichannel Slotted ALOHA System�

J. H. Sarker & A. Ghanmi

27th ICCRTS October 25-27, 2022

System Model

7/(29)

8 of 29

System Model and Performances of Conventional Multichannel Slotted ALOHA System�

27th ICCRTS October 25-27, 2022

System Background and Basics of Resultant Vector or Readiness

J. H. Sarker & A. Ghanmi

8/(29)

9 of 29

System Model and Performances of Conventional Multichannel Slotted ALOHA System�

27th ICCRTS October 25-27, 2022

Throughput Performance of Any Multichannel ALOHA System

Throughput per channel

Probability of Success

J. H. Sarker & A. Ghanmi

Input traffic per channel

9/(29)

10 of 29

System Model and Performances of Conventional Multichannel Slotted ALOHA System�

27th ICCRTS October 25-27, 2022

Resultant Vector and Normalized Readiness

J. H. Sarker & A. Ghanmi

Let throughput performance is S, delay performance is D, and threat-hazard resiliency performance is T as three vectors

The readiness is defined as the resultant vector of these three vectors

The readiness is

The normalized readiness is

10/(29)

The optimum traffic load G/M to make the throughput Q maximum will play very important role.

11 of 29

Performance Analysis of Pure and Slotted ALOHA

27th ICCRTS October 25-27, 2022

Multichannel pure ALOHA System

J. H. Sarker & A. Ghanmi

11/(29)

12 of 29

Performance Analysis of Pure and Slotted ALOHA

27th ICCRTS October 25-27, 2022

Multichannel pure ALOHA System

J. H. Sarker & A. Ghanmi

Throughput with M channels pure ALOHA system is

12/(29)

Throughput performance

After solving the equation

13 of 29

Performance Analysis of Pure and Slotted ALOHA

27th ICCRTS October 25-27, 2022

Multichannel pure ALOHA System

J. H. Sarker & A. Ghanmi

Normalized delay for pure ALOHA

For Pure ALOHA

Value of normalized delay of Pure ALOHA

Delay performance for Pure ALOHA

13/(29)

14 of 29

Performance Analysis of Pure and Slotted ALOHA

27th ICCRTS October 25-27, 2022

Multichannel pure ALOHA System

J. H. Sarker & A. Ghanmi

Threat-hazard resiliency of pure ALOHA is

For Pure ALOHA

Threat-hazard resiliency performance for Pure ALOHA is

14/(29)

15 of 29

Performance Analysis of Pure and Slotted ALOHA

27th ICCRTS October 25-27, 2022

Multichannel pure ALOHA System

J. H. Sarker & A. Ghanmi

Performance Metrics

Pure ALOHA

Through performance

Delay performance

Threat-hazard resiliency performance

Readiness

Normalized readiness

0.184

0.368

0.368

0.552

0.319

15/(29)

16 of 29

Performance Analysis of Pure and Slotted ALOHA

27th ICCRTS October 25-27, 2022

Multichannel slotted ALOHA System

J. H. Sarker & A. Ghanmi

16/(29)

17 of 29

Performance Analysis of Pure and Slotted ALOHA

27th ICCRTS October 25-27, 2022

Multichannel slotted ALOHA System

J. H. Sarker & A. Ghanmi

Throughput with M channels slotted ALOHA system is

17/(29)

After solving the equation

Throughput performance

18 of 29

Performance Analysis of Pure and Slotted ALOHA

27th ICCRTS October 25-27, 2022

Multichannel slotted ALOHA System

J. H. Sarker & A. Ghanmi

For Slotted ALOHA

Value of normalized delay of slotted ALOHA

Delay performance for slotted ALOHA

18/(29)

Normalized delay of SA is

19 of 29

Performance Analysis of Pure and Slotted ALOHA

27th ICCRTS October 25-27, 2022

Multichannel slotted ALOHA System

J. H. Sarker & A. Ghanmi

Threat-hazard resiliency is

For slotted ALOHA

Threat-hazard resiliency performance for slotted ALOHA is

19/(29)

20 of 29

Performance Analysis of Pure and Slotted ALOHA

27th ICCRTS October 25-27, 2022

Multichannel slotted ALOHA System

J. H. Sarker & A. Ghanmi

Performance Metrics

Slotted ALOHA

Through performance

Delay performance

Threat-hazard resiliency performance

Readiness

Normalized readiness

0.368

0.368

0.368

0.637

0.368

20/(29)

21 of 29

Improved Normalized Readiness with MLeFRSA

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

A random access algorithm comprises of machine learning empowered fast retransmission trials and Slotted ALOHA (MLeFRSA)

Group 1

Group 2

Total

21/(29)

22 of 29

Improved Normalized Readiness with MLeFRSA

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

Throughput with M channels MLeFRSA system is

22/(29)

After solving the equation

23 of 29

Improved Normalized Readiness with MLeFRSA

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

Normalized delay for MLeFRSA

For MLeFRSA

23/(29)

Delay performance for MLeFRSA

24 of 29

Improved Normalized Readiness with MLeFRSA

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

Threat-hazard resiliency MLeFRSA

For MLeFRSA

24/(29)

Threat-hazard resiliency performance for MLeFRSA

25 of 29

Improved Normalized Readiness with MLeFRSA

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

Performance Metrics

MLeFRSA

Through performance

Delay performance

Threat-hazard resiliency performance

Readiness

Normalized readiness

0.6004

0.6004

0.6004

1.0399

0.6004

25/(29)

26 of 29

Discussion 1/(2)

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

Performance Metrics

Pure ALOHA

Slotted ALOHA

MLeFRSA

Max. Value

Throughput performance

0.184

0.368

0.6004

1

Delay performance

0.368

0.368

0.6004

1

Threat-hazard resiliency performance

0.368

0.368

0.6004

1

Readiness

0.552

0.637

1.0399

Normalized readiness

0.319

0.368

0.6004

1

26/(29)

27 of 29

Discussion 2/(2)

27th ICCRTS October 25-27, 2022

J. H. Sarker & A. Ghanmi

Normalized readiness

0.319

0.368

0.6004

1

Performance Metrics

Pure ALOHA

Slotted ALOHA

MLeFRSA

Max. Value

Endeavour Complexity

Appropriate C2 Approach

Low

De-conflicted

Medium

Coordinated

Medium-High

Collaborated

High-Very high

Edge

Appropriate Communications Algorithm

Pure ALOHA

Slotted ALOHA

MLeFRSA

Future RA

Same network

LoRaWAN

Cost and Complexity will reduce significantly

27/(29)

Low

Medium

Medium-High

High-Very high

28 of 29

Conclusions and Future Works

J. H. Sarker & A. Ghanmi

There is no single best approach networks that support C2.

Implementation of a single ad hoc network with multiple communications algorithms reduce the cost and complexity significantly;

at the same time may fulfill the requirements of C2 Agility.

EW and Communications and information systems (CIS)

Therefore, threat-hazard resiliency performance

C2 agility and a hypothetical relationship between the appropriate communications algorithms, appropriate C2 Approaches and endeavour complex missions have also been presented in this paper.

Effects of human-in-the-loop and human-on-the-loop on C2 is one of our future research works.

27th ICCRTS October 25-27, 2022

28/(29)

29 of 29

Questions?

J. H. Sarker & A. Ghanmi

27th ICCRTS October 25-27, 2022

You can also send your question to:

dr_jahangir_h_sarker@yahoo.com

29/(29)

Thank you for your kind attention