汽车嵌入式安全：深度探讨与未来应用

需积分: 10 161 浏览量更新于2024-07-25 1 收藏 3.17MB PDF 举报

嵌入式安全在车辆中的应用是一本由Kerstin Lemke、Christof Paar和Marko Wolf三位专家合编的著作，着重探讨了信息技术（IT）安全在现代汽车行业中的关键作用。本书深入剖析了当前和未来汽车电子系统中的安全挑战，提供了53幅图表和25个表格来辅助理解，确保读者能够全面了解这一领域的最新进展。随着汽车电子化程度的不断提高，车辆正逐渐转变为复杂的移动计算平台，这使得它们成为网络安全的重要靶标。嵌入式安全在汽车中的应用包括保护关键的车载信息娱乐系统、车辆通信网络、电子控制系统（如制动、引擎控制等）免受黑客攻击。书中可能涵盖了以下主题： 1. **基础概念**：介绍了嵌入式系统的安全特性和汽车电子架构，强调了为何安全性对于防止篡改、数据泄露和潜在的物理安全威胁至关重要。 2. **威胁模型**：分析了汽车面临的各种威胁，如远程攻击、恶意软件、物理破坏以及零日漏洞利用，这些都可能导致驾驶安全问题和隐私侵犯。 3. **加密与认证**：书中可能会详细介绍如何通过加密技术（如SSL/TLS）保护通信数据，以及数字签名和身份验证机制在确保车辆通信真实性和完整性方面的作用。 4. **安全协议与标准**：涵盖了国际和行业内的安全标准，如ISO 26262（功能安全）和SAE J3061（信息安全），以及它们如何指导汽车制造商在设计和实施安全措施时遵循最佳实践。 5. **安全测试与评估**：探讨了如何对汽车电子系统进行安全审计、漏洞扫描和渗透测试，以发现并修复潜在的安全隐患。 6. **实时操作系统（RTOS）安全**：强调了RTOS在汽车安全中的关键角色，以及如何在实时性和安全性之间取得平衡。 7. **安全更新与补丁管理**：讨论了如何有效地更新车载软件，防止已知漏洞被利用，并管理长期存在的安全风险。 8. **未来趋势与挑战**：书中可能还对未来汽车安全的发展进行了预测，例如车联网（V2X）、自动驾驶车辆以及新兴技术（如区块链和人工智能）如何影响安全策略。 9. **案例研究**：通过具体的案例分析，展示了实际的嵌入式安全解决方案在汽车行业的应用实例，以供读者学习和借鉴。这本书为IT专业人士、汽车工程师和安全研究人员提供了一个深入理解汽车电子系统安全现状与未来发展趋势的宝贵资源。通过对嵌入式安全的详细研究，它旨在帮助行业建立更强健、更可靠的安全防护体系，确保驾驶者和乘客的信息安全以及车辆的正常运行。

8 Christof Paar

cell phones. Other approaches make use of biometrics, e.g, ﬁngerprint recogni-

tion. Another class simply requires active user input in order to communicate

the person’s identity to the vehicle. Again, we will need security technologies

here in order to prevent abuse. Technologies needed included identiﬁcation

techniques, biometrics and tamper resistance hardware.

Access Control for Car Data. Already today many cars are equipped

with event data recorders. This can be as simple as tachograph data, or more

advanced systems which record a wide variety of information about the car

subsystems and driving behavior. Currently, such data can usually only be

accessed via diagnosis interfaces which have to be attached physically to the

car. However, it seems likely that many vehicles will be equipped with wire-

less interfaces such as Bluetooth or GSM in the future. It becomes crucial

now to tightly control access to both technical data about the car and stored

information about driving behavior. Relevant security functions are authen-

tication and identiﬁcation protocols and communication security. The article

Security Aspects of Mobile Communication Systems deals with security topics

for wireless protocols.

Anonymity. Cars ﬁlled with IT systems oﬀer several possibilities for vi-

olating drivers’ privacy rights. The above-mentioned recording of driving be-

havior is one example. Navigation data used or requests for other location-

based services (e.g., the purchase of certain navigation data, requests for the

nearest gas station or requests for the nearest restaurant) is another exam-

ple. It is also imaginable that even traﬃc violations, e.g., driving beyond the

allowed speed limit, are recorded. These can all be serious threats in an in-

formation society and it will be crucial to prevent abuses by incorporating

technologies such as access control and anonymization.

Legal Obligations. Already today there are several regulations that dic-

tate the inclusion of IT security functions in cars. An example is Toll Collect,

the German road toll system or the European tachograph, as described in the

contribution A Review of the Digital Tachograph System in this volume. In the

future, there will be more applications which require IT security due to legal

regulations. Possible examples include emergency call systems, immobilizers

or other theft control measures, and event data recorders.

We do not claim that the listing above is complete. However, we believe

that embedded security is already an important technology for a host of di-

verse car functions, and its impact will increase in the future. In summary, it

can be claimed that IT security will play the role of an enabling technology

for numerous future car applications.

3 Embedded Security Technologies in Vehicles

3.1 Embedded Security vs. General IT Security

Since the late 1990s embedded security, sometimes also referred to as security

engineering or cryptographic engineering, has emerged as a proper subdisci-

Embedded IT Security in Automotive Application – An Emerging Area 9

pline within the security and cryptography communities. Embedded security

is often quite diﬀerent from the security problems encountered in computer

networks such as LANs or the Internet. For such classical networks there exist

established and relatively mature security solutions, e.g., ﬁrewalls, encryption

software, and intrusion detection systems. The topics with which embedded

security deals are, generally speaking, closer related to the underlying software

and hardware of the target device which is to be protected. Arguably the most

important event at which embedded security technologies are treated from a

scientiﬁc viewpoint is the CHES (Cryptographic Hardware and Embedded

Systems) Workshop series [4, 5, 6, 7, 8, 9, 10] which started in 1999.

Even though there are certainly many aspects of security that are shared

by embedded devices and general computers, there are a number of key diﬀer-

ences: First, embedded devices tend to have small processors (often 8-bit or

16-bit micro-controllers) which are limited with respect to computational ca-

pabilities, memory, and power consumption. Modern PCs, on the other hand,

are very powerful and in most cases do not limit the use of cryptographic func-

tions. Second, potential attackers of embedded systems have often access to

the target device itself, e.g., an attack of a smart card only makes sense if one

actually has physical control over the smart card. On the other hand, attacks

against traditional computer networks are almost always performed remotely.

Third, embedded systems are often relatively cheap and cost sensitive because

they often involve high-volume products which are priced competitively. Thus,

adding complex and costly security solutions is not acceptable. By comparing

typical prices (e.g., a laptop vs. an ECU) one easily notices a ratio of 1–2

orders of magnitude which, of course, limits the costs that can be spent on

security for embedded solutions.

3.2 Cryptographic Algorithms in Constraint Environments

Even though security depends on much more than just cryptographic algo-

rithms – a robust overall security design including secure protocols and or-

ganizational measures are needed as well – crypto schemes are in most cases

the atomic building blocks of a security solution. The problem in embedded

applications is that they tend to be computationally and memory constrained

due to cost reasons. (Often they are also power limited, but since automotive

applications are often powered by their own battery, low-power crypto is not

such an important topic in the car context.) It is now the task of the embed-

ded security engineer to implement secure crypto algorithms on small devices

at acceptable running times.

Crypto schemes are divided into two families: symmetric and asymmetric

algorithms. The ﬁrst group is mainly used for data encryption and message

integrity checks. Symmetric algorithms tend to run relatively fast and often

need little memory resources. There exists a wealth of established algorithms,

with the most prominent representatives being the block ciphers DES (Data

Encryption Standard) and AES (Advanced Encryption Standard.) The family

10 Christof Paar

of stream ciphers can be even more eﬃcient than block ciphers and are, thus,

sometimes preferred for embedded applications. In almost all cases it is a

wise choice to use established, proven algorithms rather than unproven or

self-developed ones. More on the state-of-the-art of symmetric algorithms will

be said in the contribution Fundamentals of Symmetric Cryptography of this

volume.

The second family of schemes, asymmetric or public-key algorithms, are

very diﬀerent. They are based on hard number theoretical problems and in-

volve complex mathematical computations with very long numbers, commonly

in the range of 160–4048 bits, depending on the algorithm and security level.

Their advantage, however, is that they oﬀer advanced functions such as dig-

ital signatures and key distribution over unsecure channels. For common au-

tomotive applications such as secure ﬂashing, public-key algorithms are of-

ten preferred. The problem here is the computational requirement of public-

key schemes. Embedded processors in the automotive domain are often only

equipped with 8-bit and 16-bit processors clocked at moderate frequencies

of, say, below 10 MHz. Running computationally expensive public-key algo-

rithms on such processors can result in unacceptably long execution times,

for instance several seconds for the generation of a digital signature. For this

reason, it is very important that a smart parameter choice together with

the latest implementation techniques are being employed. Much more details

about properties and the selection of public-key algorithms will be given in

Fundamentals of Asymmetric Cryptography of this volume.

3.3 Physical Security: Side Channel Attacks and Reverse

Engineering

A central tool for providing security are cryptographic algorithms. Both sym-

metric and asymmetric algorithms are based on the fact that the protected

device (e.g., tachograph, an ECU, or an infotainment device) is equipped with

a secret cryptographic key. “Secret” means in this context also that it can not

be read out by an attacker. If an attacker obtains knowledge of the key, the

device can usually be manipulated and/or cloned. Many of the potential at-

tackers – which includes in particular the owner and maintenance personnel

– have physical access to the device.

One family of attacks which attempt to recover the key from the device are

side channel attacks, which were ﬁrst proposed in the open literature in 1996

[11]. Side channel attacks observe the power consumption, the timing behav-

ior or the electromagnetic radiation of an embedded device. These signals are

recorded while the cryptographic algorithm with the secret key is executed.

The attacker then tries to extract the key by means of signal processing tech-

niques. Side channel attacks are a serious threat in the real world unless special

countermeasures have been implemented. Much more about side channels will

be said in the contribution Embedded Cryptography: Side Channel Attacks in

this volume.

Embedded IT Security in Automotive Application – An Emerging Area 11

A related family of attacks are fault injection attacks, sometimes referred

to as active side channel attacks. Fault injection attacks force the device to

malfunction, for instance by spikes in the power supply, through overclocking,

or through overheating of the embedded device. The goal is often to create

an incorrect output of the cryptographic algorithm which leaks information

about the key used.

Quite diﬀerent from side channel and fault injection attacks are reverse

engineering attacks. The goal here is to read the cryptographic keys directly

from the RAM, EEPROM, FlashROM, or ROM of the embedded device.

Unlike classical reverse engineering of code it is in this context sometimes

suﬃcient to recover a single cryptographic key for a successful attack, which

is often only 16 bytes long or less. Of course, there is tamper-resistant memory

available but it is for automotive systems often not available for cost or legacy

reasons. Case studies about tamper resistance in the real world are described

in [2]. A more detailed treatment is given in the article Embedded Security:

Physical Protection against Tampering Attacks in this volume.

3.4 Digital Rights Management (DRM)

DRM has become a very important technology for applications such as audio

and ﬁlm distribution over the Internet. DRM systems can enforce rules such as

the time period for which access to a music ﬁle is granted or to which device a

digital movie is allowed to be copied. It is perhaps a bit surprising that DRM

should become important for vehicles as well. However, as soon as digital data

used for car applications represent ﬁnancial values, e.g., ﬂash software, digital

location-based services or entertainment content, DRM will be the technology

that enforces the envisioned use of the data. DRM technologies are required to

prevent the customer from unauthorized copying or an unauthorized extension

of the usage period of the content.

In order to realize a proper DRM platform in a vehicle, we need trusted

computing functions which in turn are based on physical secure components

such as secure memory, true random number generators and cryptographic

algorithms.

3.5 Further Topics

The topics discussed above are certainly not comprehensive. However, they

form a core of embedded security technologies that are relevant for most se-

curity solutions in cars. Topics such as mobile security are also treated in this

volume.

12 Christof Paar

4 Conclusion: Challenges and Opportunities for the

Automotive IT Community

In summary it can be stated that embedded IT security in cars:

1. protects against manipulations by outsiders, owners and maintenance per-

sonnel,

2. increases the reliability of a system,

3. enables new IT-based business models.

As sketched above, there are several diﬃculties to overcome in order to de-

velop strong embedded security solutions. We would like to give an outlook on

the future of IT security in cars in the form of the following recommendations

and conclusions:

• IT security will be a necessary requirement for many future automotive

applications.

• Security will allow a multitude of new IT-based business models, e.g.,

location-based services or fee-based ﬂashing. For such systems, security

will be an enabling technology.

• Security will be integrated invisibly in embedded devices. Embedded secu-

rity technologies will be a ﬁeld in which manufacturers and part suppliers

need to develop expertise.

• Security solutions have to be designed extremely carefully. A single “minor”

ﬂaw in the system design can render the entire solution unsecure. This is

quite diﬀerent from engineering most other technical systems: a single

non-optimum component usually does not invalidate the entire system.

An example is the Content Scrambling System (CSS) for DVD content

protection, which was broken easily once it was reverse engineered.

• Embedded security in vehicles has to deal with very speciﬁc boundary

conditions: computationally and memory constrained processors, tight cost

requirements, physical security.

• The multi-player manufacturing chain for modern vehicles (OEM and pos-

sibly several layers of suppliers) can have implications for the security de-

sign. It is, for instance, relevant who designs a security architecture and,

most importantly, who has control over the cryptographic keys.

• Merging the automotive IT and the embedded security community will

allow many new applications. However, there are also several challenges:

security and cryptography has historically been a ﬁeld dominated by the-

oreticians, whereas the automotive IT is usually done by engineers. The

culture in those two communities is quite diﬀerent at times, and both

sides have to put eﬀort into understanding each other’s way of thinking

and communicating.

剩余270页未读，继续阅读

benlyhong

粉丝: 0
资源: 5

汽车嵌入式安全：深度探讨与未来应用

嵌入式系统在车辆综合电子信息系统中的 应用研究.pdf

嵌入式系统在车辆综合电子信息系统中的应用研究.pdf

嵌入式系统有哪些应用领域

嵌入式系统有哪些应用领域？

嵌入式视频监控系统应用背景

写一篇有关嵌入式实际应用的结论

嵌入式最火的10个行业

嵌入式在智能交通安防领域的应用

嵌入式中uC/OS-II的应用

阐述嵌入式部署目前所达到的性能，能够适应和符合哪些实际汽车应用，或者是未来可能怎么应用，或者是阐述在最终实车应用的路上有推进作用

最新资源

嵌入式系统在车辆综合电子信息系统中的应用研究.pdf