Autonomous Security - armkeil.blob.core.windows.net · 2) Securing IoT: OEM responsibility or end user’s dilemma? 3) Autonomous Security’s use of hardware Root of Trust & machine

Autonomous SecurityHardware-Based Security Model & Trust Stack

Jay Fallah CTO NXMArm TechCon 2019, Booth #931

Outline

1) About us

2) Securing IoT: OEM responsibility or end user’s dilemma?

3) Autonomous Security’s use of hardware Root of Trust & machine identity

4) Cyber Physical Security vs. Cyber Security

5) Evolution of security: personal computing, to mobile, to IoT & beyond

6) From centralized systems to distributed & decentralized models

7) Questions

© 2019 NXM Labs, Inc. - Propietary & Confidential 2

NXM - Next Generation Machines

Founded in 2016

Developing Autonomous IoT Security platform & Data Integrity trust-stack

‘Chip-to-Cloud’ solution leverages PSA & trusted enclave

Frost & Sullivan 2019 NA Visionary Innovation Leadership Award for IoT Security


Catch MEIF

YOU Can


About a decade ago an FBI agent explained:“There are 2 types of people”


Those who have experienced

a cyber-threat

And Those who have not

A few years later the same FBI agent said:“There are two groups of people”


Those who have experienced

a cyber-threat

And Those who do not know

that they have

“Amateurs HACK DevicesProfessionals HACK People”

Criminals chose targets based on least effort for the highest reward

Cybercrime damages has surpassed all illegal drugs crime

Phishing attacks of the past have graduated to whaling


Real People NOT Actors


Do you feel secure? Are you secure?

Are you a human firewall?

Is feeling secure the same as being secure?

Why do we all need to be vigilant about security?

Isn’t there a better way than username, passwords & settings?

Would it be possible to secure the tsunami of future IoT devices using password?


Are we making smart devices more secure?


Replace “Smart” with “Hackable”Replace “Connected” with “Exposed”

SMART CITY

SMART DEVICES

CONNECTED DEVICES

SMART CONNECTED VEHICLES


HACKABLE CITY

HACKABLE DEVICES

EXPOSED DEVICES

HACKABLE EXPOSED CARS

=

=

=

=

Feeling Secure ≠ Being SecureThey relate to each other by a model


Vulnerable: Not secure & not feeling secure

False sense of security: Not secure but feeling secure

Paranoid: Secure but feeling insecure

Secure: Being secure & feeling secure

Model

Feeling Secure

Being Secure

IoT Security ≠ Internet SecurityAutonomous Security ≠ Cyber Security


IoT proliferation is based on use of analytics & big data

Autonomous Security is a Cyber Physical Security

Its purpose is securing Edge Computing devices

Utilizing hardware to anchor cybersecurity primitives in TEE

“Innovation”


The action or process of innovating. The invention or application of methods, ideas or devices. From Latin, “innovare”, to renew or make new.

Autonomous Security is the modelfor “Edge Cyber Physical” security


Value Chain: ARM, Chip, Module, OEM, User (B2B2C)

Life Cycle: Make, Ship, Sell, Provision, Use, Discard

Target User: OEM (integrating security in final design)

Autonomous: Provisioned upon first network connection

Chip Designer

Chip Maker

Module Maker

OEM

User

Money is the currency of transactionsTrust is the currency of interactions


Evolution of Trust: Local, Institutional, Distributed

Autonomous Security: Distributed Trust Stack

Securing: Firmware, Software, Data, Communications

Based on: Hardware & Trusted Firmware

Known UnknownTrust

Uncertainty

3. other user

2. platform

1. idea

THE TRUST STACK

Known Unknown

Autonomous SecurityNeighborhood Watch of/for/by Devices


Security interactions separated from data interactions

Configurations defining data context & data path

All data signed and encrypted at rest & inflight

Single source of truth in a collaborative digital ledger

Automated Configuration ManagementPre-definition rather than post-learning


Data integrity & privacy at the edge

Network segmentation based on device identity

Can you trust Big Data if you can not trust Little Data?

Should IT & the vendor collaborate on device security

Delegating system & data threats to the hardware


Trusted Firmware & PSA CertificationSecurity requires friction & delimitation


OEM should be responsible as the integrator for security

Security vendors target IT departments as purchasers

Operational technology must have a security component

Chip designers are too far from device security

Evolve security from a commodity to a common good

Security journey of an IoT must start in the field

Manufacture Build

Shipping

Wholesale Retail

Provisioning

Ownership changes

Disposal

Digital ledger tracking security

Warranty

Late

-Bin

ding

Late-binding & life-time Warranty


Tracked By Serial number

Autonomous SecurityConfigurationMachine ID

PKI

Ledger as IAM & Key Exchange Management


Globally unique device identity (Late-Binding, TRNG).

PKI created & private key stored in the trusted zone.

Identity & public key provisioned on the blockchain.

Identity tracks the device security posture in the ledger.

What is stored on the Ledger? Do we really need a blockchain?Why can’t databases do the same thing?


Security needs a distributed & decentralized model

Variables are Identity, public key & a JSON string

Blockchain is not just a database in this implementation

Smart Contracts allow customized device programming

Smart Contracts could be delimited by resources

Who owns the device? And who owns the blockchain?


Autonomous devices means no human ownership

One admin device & 0 to ∞ users provisioned in field

4 possible implementation of permissioned blockchain

Ethereum & Hyperledger blockchain microservices

Security & only security is handled by the ledger

Autonomous Security for next generation machines

IOT

Data sent to cloud

High Latency

Not Scalable

High Bandwidth Needs

M2M

Machine to Machine interaction

Realtime

Scalable

Edge processing of data


Two views on any IoT infrastructure

SYSTEM VIEW

Things, IoT devices, machines

Gateways, networking hubs, routers

Network services, connection

Cloud services, APIs, IoT Frameworks

Device management, control

BUSINESS VIEW

IoT platform, big data, analysis

Connectivity, device management

Business model, business transactions

Services offered, sectors served

Applications, customer interactions


Edge devices & IoT requirements


Requirement: Power, Connectivity (IP at the edge)

Abilities: Sensing, actuating or both mostly headless

Trusted: Enclave, Storage, Application, Memory, TRNG

Categories: Automotive, 5G, IoT (Health, Smart City, …)

MACHINES THAT WORK

INDEPENDENTLY

PEERS IN NETWORK

AUTOMATED

RESILIENT

Cyber Physical Root of Trust


Trusted Software

Trusted Hardware

CryptoSecureSystem

SecureStorage TRNG

Trusted

Non-trusted


CUPSControl

and User Plane

Separation

CUPS in 5G LTE as inspiration


Layer 2 Layer 3

FDD

TDD

Layer 1

25.3

21 –

MA

C

25.3

22 –

RLC

BM

CP

DC

P

Physical Layer

RR

C

CP

UP

Control Plane

User PlaneBroadcast

Packet Switched

Circuit Switched

Using blockchain as overlay layer for IoT control


Layer 2 Layer 3

RF

Wire

Layer 1

MA

C

IP

Physical Layer

over

lay

CP

UP

Control Plane

User Plane

MQTT Pub-Sub

DLT

Discovery & visibility in the NetworkFunctional edge network segmentation


Single security context based on device identity

Multiple data/service context for each physical device

Device mission specification by OEM at configuration

Single admin, multi-user & life time device identity

Trusting the device starts with Root of Trust (RoT)PKI involved in creating as well as extending the RoT


Autonomous cyber physical devices must safeguard their RoT

Machine identity is essential to implementing an IoT security framework

Firmware, application, data & communication should all be trustworthy

A trust stack allows devices to trust each other based on rules on engagement

IAM & key management should be event based, auditable & distributed

Mission-based edge devices form the first line of defense against intruders

Summary1: Catch me if you can!

I can only catch the misbehaving physical devices that I can identify

Device should be able to do secure boot & secure update over the air

The evolution of computing (PC, Mobile, IoT) each need their own security journey

We need more than IT people to deal with security issues of a trillion IoTs

Monitoring device behavior makes catching a rogue actor easier


Summary2: Human firewall is not working

You only remember the combination of a lock that you open on daily basis

Should OEM be responsible for the security of the device or should the owner

Swimming with the sharks only works when you can keep track of the sharks

Autonomous vehicles need Autonomous Security, humans not required

Security paradigm should include privacy & data integrity

Professionals hack people, let their devices defend them!


Summary3: Realtime or near real time remedy

Getting back to state of grace means repairing the damage & replacing RoT

All authorized access recorded in a decentralized & distributed ledger

Ledger providing actionable third party for peer-to-peer interactions

The remedy begins when the numbers are different

Networking as well as computing needs to be controlled & delimited

IoTs are made with predetermined missions in mind & can be configured as such


QuestionsJay Fallah, CTO NXM Labs

[email protected]

+1-647-927-9990

Visit us at Booth #931 for a Demo!


mailto:[email protected]

Documents

Autonomous Security - armkeil.blob.core.windows.net · 2) Securing IoT: OEM responsibility or end user’s dilemma? 3) Autonomous Security’s use of hardware Root of Trust & machine