Demystifying Intelligent Multimode Security Systems: An Edge-to-Cloud Cybersecurity Solutions Guide

Table of Contents
About the Authors
About the Technical Reviewer
About the Intel Reviewers
Acknowledgements
Legal Notices and Disclaimers
Abstract
Chapter 1: Introduction and Overview
Why You Should Read This Introduction and What to Expect
…Because That’s Where the Money Is
Cogito Ergo (Multiply and) Sum – Artificial Intelligence
What This Means for You As a Security and Safety Professional
Every Journey Begins with a Single Step – Maya Angelou
Chapter 2: IMSS System Level View
Summary
History of Intelligent Multi-modal Security System Solutions
Video 1.0 – Analog Video Technology
Video 2.0 – IP-Connected Cameras and Video Recorders
Current Intelligent Multi-modal Security Systems Solutions
Video 3.0 – Intelligent Cameras and Video Recorders
Bandwidth and Connectivity
Cost/Power/Performance
Ease of Development, Deployment, and Scaling
Next-Generation Intelligent Multi-modal Security Systems Solutions
Impact of Memory and Compute Improvements
Design for Privacy
Personal data
Processing
Protection
Protection Security guidelines
Principle IMSS System Components
IMSS System View
Smart IP Camera
Network Video Recorder with Analytics
Compute resources – General to Specialized, Key Performance Indicators (KPIs)
Edge Server
Operations Data Center Server
End-to-End Security
Cost Overheads for Security
Confidentiality, Integrity, Availability
Secure Data Storage
Conclusion
Chapter 3: Architecting and E2E IMSS Pipeline
What Does It Take?
IMSS Data Pipeline Terminology
Defining the Data Pipeline – Key Concepts
Desired Actions and Outcomes
Three Basic Tasks – Storage, Display, and Analytics
Basic Datatypes and Relationships – Sensed Data, Algorithms, and Metadata
Evolution of IMSS Systems, or a Brief History of Crime
IMSS 1.0 In the Beginning, There Was Analog…
IMSS 2.0 …And Then There Was Digital…
IMSS 3.0 …Better Together – Network Effects…
Breaking Up Is Hard to Do…Packets Everywhere…
Learning to Share…
Hook Me Up…Let’s Get Together
Data Rich, Information Sparse…
IMSS 4.0…If I Only Had a Brain…
Classical CV Techniques – Algorithms First, Then Test against Data
Deep Learning – Data First, Then Create Algorithms
Convolutional Neural Networks and Filters…Go Forth and Multiply…and Multiply
Teaching a Network to Learn…
Types of Neural Networks: Detection and Classification
A Pragmatic Approach to Deep Learning …Key Performance Indicators (KPIs)
One Size Doesn’t Fit All…
IMSS 4.0: From Data to Information
Information Rich, Target Rich…
Task Graph – Describing the Use Case/Workload – Overview
Sensors and Cameras – Sampling the Scene in Space, Time, and Spectra
Converting Sampled Data to Video
Transporting Data – Getting Safely from Point A to Point B
NVR/Video Analytic Nodes – Making Sense of The World
Storing Data – Data at Rest
Converting Reconstructed Data to Information – Inferencing and Classification
Humans Consumption – Display
Machine Consumption – Algorithms, Databases, and Metadata
Video Analytic Accelerators – Optimized Analytics
Conclusions and Summary
Chapter 4: Securing the IMSS Assets
Why Should You Think About Threats?
Summary
Threat Modeling
Threat Modeling Terminology
Threat Taxonomy
Basic Types of Software Threats
Basic Types of Hardware Threats
Insider, Supply Chain, and Authorized Agent Threats
Side Channel Attack Threats
Threat Analysis Methods
Basic Concepts
Common Criteria for Information Technology Security Evaluation
STRIDE
IMSS Assets
Value of Assets
Foundational Assets
Data Assets
Application Assets
Threats
Attackonomics
Current Threats
Vulnerabilities
Malware
Trends and Emerging Threats
Designing a Secure Platform Using Intel Technologies
Root of Trust (RoT)
Secure IDs
Provisioning the Device – FIDO Onboarding
Secure Boot – Chain of Trust
Securing Keys and Data
Cryptographic Keys
Data in Flight
Data at Rest
Trusted or Isolated Execution
Defense in Depth
OpenVINO Security Add-on
Secure Development Lifecycle (SDL)
Support and Response
Summary
Chapter 5: Machine Learning Security and Trustworthiness
Usage of Machine Learning in IMSS
Challenges and Risks
Policy and Standards
Regulatory Landscape
IoT Security Baseline
Privacy Compliance
GDPR and Emerging EU AI Regulation
AI/ML Trustworthiness
Trustworthiness Journey
AI Model and Data Provenance
AI Risk Management
IMSS with ML Protection
Stakeholders and Assets
Threats
Threats for the Training Process
Threats for the Inferencing Process
Training at the Edge
ML-based IMSS Protection and Trustworthiness Framework
Foundational Device Security
Workload Protection
IP Protection
OpenVINO™ Model Protection Security
Data Protection, Privacy, and Provenance
Federated Learning
Homomorphic Encryption
Data Provenance
Trustworthiness
Chapter 6: Sample Applications
Putting It All Together – What Does It Take to Build a System?
Resource Graph
Crawl – Starting Small – Workstation: An SMB IMSS System
SMB System Assets and Threats
Using Information Security Techniques to Address These Threats to an SMB IMSS System
Walk – Let’s Get Enterprising – Edge Server: Critical Infrastructure
Edge Server Critical Infrastructure System Assets and Threats
Using Information Security Techniques to Address the Threats to an Edge Critical Infrastructure System
Run – Forecast: Partly Cloudy – Data Center: Building Blocks for a Smart City
Smart City Data Center System Assets and Threats
Using Information Security Techniques to Address the Threats to a Smart City System
Chapter 7: Vision – Looking to the Future
The Evolution of Intelligent Multimodal Security Systems
Intelligence at the edge
Multimodal
Mobility
Threats
Defenses
Trust
Privacy
What Should You Do?
Chapter 8: As We Go to Press
Growth of IMSS
Cybersecurity General
Technology
Artificial Intelligence and Machine Learning
Regulations
Standards
Final Exhortation
Index