Analogue, Mixed Signal, LSI


Hardware security ICs offer large security returns at a low cost

4 September 2013 Analogue, Mixed Signal, LSI

In the face of security vulnerabilities in health and safety applications, governments and industry leaders have focused on security weaknesses in the design and protection of electronic devices. This article explains how using hardware security ICs dramatically reduces the risk of unauthorised access to vital devices, peripherals and computer systems at a low cost and with minimal impact on designs.

Vulnerabilities to attacks in critical systems are driving growing awareness for the need for improved security methods. Although the industry has relied largely on software-based cybersecurity methods, hardware-based methods are quickly gaining recognition as uniquely capable of delivering strong protection and providing a basis of trust well worth their incremental cost.

The emergence of cost-effective silicon solutions enables designers to harden designs—dramatically reducing the risk of unauthorised access to embedded devices, peripherals and systems, with minimal impact on overall cost.

As an example, consider a cardiac patient resting comfortably at home as a state-of-the-art wireless pacemaker delivers a steady stream of pulses to ensure synchronised heart muscle contractions. Unknown to him, a team of opportunistic hackers have stumbled onto his pacemaker system’s IP ­address and unleashed a variety of penetration tools at it. The pacemaker keeps delivering its prescribed signals, protected behind layers of security that defeat each attack, until the hackers give up and continue their scan for easier targets. The patient is never aware of the attempt as he continues to rest peacefully.

Although this scenario is imaginary, it lies well within the realm of possibility. In fact, recent alerts by the USA’s FDA (Food and Drug Administration) concerning the vulnerability of medical equipment have put a sharp focus on the need for trusted systems able to provide a secure foundation for any connected medical device.

Reports of similar weakness in today’s electronically controlled automobiles and even wireless-enabled traffic systems have made it clear that critical systems for health and safety are vulnerable to ‘bad actors’. Paradoxically, given the potentially sizeable impacts to health and safety associated with a successful cyber attack, the cost of including a reasonable level of trust and security in a design is negligible.

For decades, trusted system protocols – such as computer systems and buildings – have relied on layered protection of resources designed to limit access only to authorised users, software processes or other hardware devices. To permit access, the protected resource needs to authenticate, or verify, the identity of the requesting entity.

While authentication methods for ­access to secure facilities rely on multiple authentication factors (including access cards, access codes and biometric measurements), authentication methods for access to individual computer systems and devices have lagged.

Even today, computer system security relies largely on simple IDs and passwords. Yet IDs and passwords can be compromised, exposing systems to viruses and more insidious malware able to sit quietly and steal confidential assets over a period of time.

In fact, in an alert to manufacturers, the FDA identified weak password security as a key vulnerability, due to “uncontrolled distribution of passwords, disabled passwords, hard-coded passwords for software intended for privileged device access (e.g., to administrative, technical and maintenance personnel).”

Indeed, devices touching any aspect of health and safety require implementation of authentication methods that are much stronger than traditional IDs and passwords familiar to computer users – and to computer attackers.

Effective authentication

A more effective approach for authentication calls for the host system to generate a random challenge. For instance, instead of the typical challenge – “What is your password?” – the challenge might be a string of seemingly random characters.

In turn, the requesting entity issues an encoded response that includes a message authentication code (MAC), which is calculated using an algorithm that comprehends not only the requesting entity’s internal data and secret, but also the specific random challenge received from the host. Then, the host compares the received MAC response with the expected response to verify that it is dealing with a recognised entity (Figure 1).

Figure 1. An effective challenge-response authentication method builds a valid response based on a random challenge to verify the identity of an authorised accessory.
Figure 1. An effective challenge-response authentication method builds a valid response based on a random challenge to verify the identity of an authorised accessory.

With this more robust challenge-response authentication approach, a wireless pacemaker, for example, can confirm that it has received a valid MAC, ensuring that it is communicating with an authorised host before altering its pulse rate. In automobiles, trucks or heavy machinery, the vehicle’s electronic controller unit can respond to apparent emergencies using data streams from trusted embedded devices and peripherals that supply valid MACs.

In the past, companies looking to implement better security measures were forced to choose between robust but expensive hardware such as standalone crypto units or accept the limitations of software-only solutions. A software implementation of a sophisticated challenge-response method may add significant load to a host pro­cessor, potentially compromising security responsiveness and overall performance of the host system.

Beyond those operational concerns, a software approach remains a weak link in any systems-security architecture. Maintained in normal system memory, secrets remain vulnerable to discovery and modification, while the algorithms themselves run on more general-purpose hardware unable to ensure complete randomness of challenges or even their penetration. The combination potentially exposes the application to attack through a variety of technological and social avenues that continue to be exploited successfully by bad actors.

Specialised ICs that implement security methods offer multiple benefits over software-based security. Specialised crypto chips free host MCUs from the processing load associated with computationally demanding crypto algorithms. Furthermore, these security ICs reduce the points of entry for attackers and provide secure storage for data such as the keys and cryptographic parameters that represent the shared secret.

By protecting confidential crypto data and algorithms behind layers of protection, these devices are able to counter all threats more ­difficult to manage in software-based security.

At a more fundamental level, secure silicon offers the basis for a root of trust that allows engineers to construct higher-level applications with reasonable assurance that the underlying foundation of algorithms and data remains secure. In fact, as discussed later, creating this root of trust in silicon ultimately depends not only on securing the silicon device itself, but also on securing the manufacturing and distribution channel starting at its source.

Secure ICs

Maxim Integrated’s DeepCover secure authenticators leverage unique physical security mechanisms. These authentication devices protect sensitive data under ­multiple layers of advanced physical security. Attackers face a device using defence-in-depth security that raises the threshold for the cost and timeliness of a successful attack beyond the perceived value of penetration.

Just as important, these hardware security ICs simplify integration. The design simplicity of the 1-Wire interface used in these low-cost devices facilitates their use in a broad array of applications. In a typical application, engineers need only add a pull-up resistor to connect a spare I/O port of an MCU to a DeepCover authentication device such as the DS28E15 (Figure 2).

Figure 2. The addition of a pull-up resistor can connect a spare I/O port of an MCU to a DeepCover authentication device.
Figure 2. The addition of a pull-up resistor can connect a spare I/O port of an MCU to a DeepCover authentication device.

Consequently, designers can easily implement secure system designs supporting ­multiple peripherals, each authenticated through a dedicated DeepCover authentication IC. Here, a DS2465 DeepCover IC serves as the 1-Wire master, handling line driving and protocol conversion between the I²C master and any attached 1-Wire slave authentication ICs.

Within an application, individual members of the DeepCover authentication family support FIPS 180-3 Secure Hash Algorithm SHA-256 and FIPS 186 public-key based Elliptic Curve Digital Signature Algorithm (ECDSA) cryptography.

For SHA-256 applications, the DS28E15, DS28E22 and DS28E25 combine an SHA-256 engine and security features with 512-bit, 2 Kb and 4 Kb user EEPROM, respectively. For ECDSA designs, the DS28E35 provides 1 Kb EEPROM along with its ECDSA engine and associated security features.

Among the security features supported across all members of the DeepCover authentication series, each device combines its user-programmable EEPROM array available for application data with protected memory areas for storing both the read-protected secret associated with the device’s supported crypto algorithm and the data parameters. In addition, each device includes a guaranteed unique 64-bit ROM identification number, which serves as input for crypto operations and provides a universally unique serial number for applications.

Figure 3. Incorporate a 1-Wire master device such as the DS2465 to handle 1-Wire line driving and protocols on behalf of an MCU.
Figure 3. Incorporate a 1-Wire master device such as the DS2465 to handle 1-Wire line driving and protocols on behalf of an MCU.

As mentioned earlier, establishing a root of trust for an IC begins at the factory and through the distribution channel. If a device’s internally stored secret were readable or modifiable by unauthorised agents, the security of systems built with that device would be immediately compromised.

Besides operational security measures, DeepCover authentication ICs enable secrets to be read-protected from unauthorised access, and write-protected to prevent any modification of the secret. Beyond this protection, the creation of the secret itself can follow recommended security principles of compartmentalisation: rather than creating the complete secret at one location, the secret can be built in stages.

As the device passes through different locations in the supply chain, it can itself generate another portion of the secret. As the authentication IC itself becomes the root of trust for the end product, the same process may apply to the finished product. The complete secret for any individual device would thus remain effectively unknown and beyond the reach of any individual.

For more information contact CST Electronics, +27 (0)11 608 0070.



Credit(s)



Share this article:
Share via emailShare via LinkedInPrint this page

Further reading:

Full sensor to cloud solution
CST Electronics Telecoms, Datacoms, Wireless, IoT
NeoCortec has demonstrated the seamless and rapid development of full sensor-to-cloud solutions using NeoMesh Click boards from MikroE and the IoTConnect cloud solution from Avnet.

Read more...
Banana Pi with NPU
CST Electronics AI & ML
The latest Banana Pi SBC, the BPI-M7, is powered by Rockchip’s latest flagship RK3588 octa-core 64-bit processor, with a maximum frequency of 2,4 GHz.

Read more...
Hardened-grade network switches
CST Electronics Computer/Embedded Technology
Lantronix’s hardened switches provide Layer 2 or Layer 3 networking, and are available as Power-over-Ethernet (PoE) or Power-over-Ethernet Plus (PoE+).

Read more...
Energy measurement IC
CST Electronics Power Electronics / Power Management
The SY7T501 and SY7T502 are Silergy’s newest energy measurement ICs for cost-sensitive IoT applications such as smart-plugs, dimmers, power-strips, intelligent lighting, and other points of load.

Read more...
Easy wireless mesh networking
CST Electronics Telecoms, Datacoms, Wireless, IoT
NeoCortec has announced three new NeoMesh Click boards that are ideal for wireless mesh networking, eliminating the need for engineers to create any development PCB hardware themselves.

Read more...
LoRa transceiver module
CST Electronics Telecoms, Datacoms, Wireless, IoT
The RYLR896 transceiver module from REYAX features the LoRa long range modem that provides ultra-long range spread spectrum communication.

Read more...
High-performance antenna solution
CST Electronics Telecoms, Datacoms, Wireless, IoT
The Spitfire Series of antennas delivers exceptional performance with its dipole design, ground plane independence, and easy wall-mount versatility for both indoor and outdoor environments.

Read more...
SMD antenna for GNSS L1 band
CST Electronics Telecoms, Datacoms, Wireless, IoT
Antenova has recently launched the Admotus SMD antenna, a ceramic antenna suitable for all GNSS positioning applications in the L1 band covering 1559 to 1609 MHz.

Read more...
High-speed PIN diode
Altron Arrow Analogue, Mixed Signal, LSI
Vishay’s new high-speed Silicon PIN diode is able to detect both visible and near infra-red radiation over a wide spectrum range from 350 to 1100 nm.

Read more...
Analogue front end for sensor measurements
Electrocomp Analogue, Mixed Signal, LSI
The NJU9103 AFE from Nisshinbo is a tiny analogue front end, with a 16-bit resolution ADC and up to 512 x signal amplification from the programmable gain amplifier.

Read more...