The digital revolution is in full swing with businesses and countries modelling how new technology-driven efficiencies can be transformative. Better healthcare, more resilient infrastructure, increased innovation, efficiency, profit margins and customer experiences – the impacts will be broad and exciting. But as the information age makes the shift from back-end processes to critical frontline business and public operations, there is an increasing focus on how to move from vision to reality. What are the fundamental building blocks that will make a global IoT real and secure?
A new target for villains
Criminality has changed, and in the future I expect James Bond to be chasing around far more hackers. Let’s hope there’s still the opportunity for car chases though, as bursting through the door of number 10 Acacia Avenue to confront a genius 14-year-old writing ransomware code isn’t quite as visually appealing. The truth is that the nature of the threat from criminals is changing – becoming increasingly invisible, menacing and widespread, and the attack surface is broadening by the day.
It doesn’t take much to think of a future trillion-dollar business that will rely on the integrity of its data. We have them already with the big data providers, but in future there will many more. Take the simple example of an energy provider: If the company and its customers can’t trust energy meter data, how can the business succeed? One hack could, in theory, kill the company.
Even a small scale disruption of a company’s infrastructure could compromise the integrity and trust in its data. It presents the risk of customer loss, legal action and company shareholder revolt. There are other implications too: being able to understand if a house is using more energy might signal the occupants are out. Yes, these are all theoretical risks that most companies would say their systems would counter. But how many times have we seen companies explaining data has been stolen and they haven’t yet worked out how? An IoT world is one with possibilities for all, including those on the wrong side of the law. Because of these universal risks, we mustthink differently about how IoT systems must be built, and how they’ll be secured.
Any device, any data, any cloud
Smart businesses and cities are complex environments, with a mass of connected devices that may be mains-connected, or ultra-low power and running on tiny batteries or a range of possibilities in between. As a result, the IoT platform managing them needs to be flexible. We think of this as an “any device, any data, any cloud” approach, or a “triple A platform.”
The idea of a triple A platform comes with the expectation of trust delivered through robust security from chip to cloud. It’s akin to a financial product that is rated for risk – triple A is a guarantee of quality, trust and a standard for investment confidence. It’s this level of assurance that we think is required for the IoT.
So, what will it take to truly deliver a “triple A” rated IoT platform? Not surprisingly, it starts with fundamentals and the device foundation. Last year Arm launched the Platform Security Architecture (PSA) as a set of principles to provide standardization over howsecure IoT devices are designed. The approach analyzes the threat in the IoT application and identifies the key features that need to be implemented in the device, hardware and software. We’re building on that this year and ensuring we provide a portfolio of IP, software and tools that enable our silicon partners to design securely from the ground up for a wide range of embedded devices, from the simplest cost constrained, through mainstream to the most advanced rich nodes and gateways. This portfolio of IP includes security IP, system components, processors and sub-systems that provide a secure foundation for chips that can be customized for a range of applications, while maintaining compliance with the PSA.
‘Triple A’ trust maintained after the device is deployed
The PSA also dictates that a device can only be secured through its lifetime if management is put in place. This means tackling three key elements: data ingest, device connectivity and device management.
Let’s start with the data ingest challenge. A complete solution must identify the source of data and translate the data into a form where the owner can operate on the data and understand the level of trust to apply to the incoming data. Securing this layer is vital, as it is where business insights are derived and actioned.Next is device management, a critical element for infrastructure owners in particular because of the need to securely manage device connectivity. The key requirement is to enable global device deployments across different communications technologies and securely distribute permissions so any device can access any operator network.
Achieving this universal connectivity (“any device”) comes with the need to provide secure onboarding and lifecycle management to mitigate device vulnerabilities from silicon birth to silicon scrapyard, and to do that for any capability of device. This need extends from secure endpoints (with built-in security architecture for authentication), to secure and resilient networking (providing access to multiple providers while maintaining user traffic and device separation), secure device management (via over-the-air updates and device-data communication) and secure data management for analytics.
The final step requires cloud services so device threats can be identified in the field: isolating, maintaining and updating a device’s firmware as vulnerabilities and attacks are identified,and so maintaining the integrity of trusted firmware. Arm recognizes that these elements are typically handled by separate organizations and that it can be challenging to put them together in a way that maintains continuity and trust. But only a platform that addresses all the elements together.
Forging the links of a trust chain
Traceability is another challenge in a complex value chain. How can the enterprise be aware of the interactions that have happened to the device from which they have just received data? Confidence in any data-driven insight is only as strong as the trust businesses and consumers can place in it. So, the trust chain for each action from chip to cloud and services, needs to be able to validate each transaction and guarantee traceability of the secure transaction. Also, a solution must offer the right to be forgotten when requested by data owners. Any IoT-driven business operating at any scale needs to manage all these elements in the simplest way possible. That requires a single-pane-of-glass view that encompasses secure and reliable connectivity using customer-based traffic/device segregation, secure public IP functionality, and N+1 resilience. They also need a single view for customers’ service billing and analytics, and the ability to tie into any connectivity network through an enterprise client management platform (“any cloud”). All of this is now covered by Arm, through our hardware ecosystem and the Arm Pelion IoT Platform.
Summary
Returning to Agent 007 and where his efforts may be focused in future. The world’s most famous spy has already faced a fictional cybercriminal in Skyfall and real-life hackers when a script for Spectre was stolen in 2014. Maybe now he’s on the trail of the cyberattackers who exploited a security flaw in Facebook’s systems to take control of users accounts in September this year. The company said 90 million people should take action as a result.
So, the threat is real and growing, and while we’re still in only a relatively connected world, we can’t wait for a trillion IoT devices to go live before fixing the holes we can see today.
