Our experiences with personal computers and the Internet have clearly identified information security as a paramount challenge. Embedded systems, which are used pervasively in our lives, now contain our sensitive personal data, identity, and even our purchasing power, and perform several safety-critical functions. Some examples include mobile phones, MP3 players, automotive electronics, medi-cal appliances, and ubiquitous devices such as sensors and RFID tags. Unless embedded system security is adequately addressed, it will become a concern that impedes the adoption and usage of many embedded system products, applications, and services.
Several technologies have been developed to address information security (cryptography, secure communication protocols, anti-virus tools, firewalls, intrusion detection, and so on), which can be adapted to embedded systems. These technologies can be referred to as "functional" security meas-ures, since they usually specify functions that must be added to the target system without any consid-eration of how they are embodied in hardware or software.
However, they are hardly sufficient to ensure the security of embedded systems in practice. Most real security attacks do not directly take on the theoretical strength of cryptographic algorithms; instead, they target weaknesses in a system's "implementation". Moreover, embedded-system designers have to cope with security as yet another requirement, in addition to performance, power, cost, etc.
We will present an introduction to embedded system security challenges, and argue that ef-fective security solutions can be realized only if they are built-in at various stages of the design process (architecture, HW design, and SW development). The objectives of secure embedded system design will be defined from the designer's perspective as addressing various "gaps" such as
1. the assurance gap, which refers to the gap between functional security measures and truly secure implementations,
2. the security processing gap, which arises due to the processing requirements of the additional computations that must be performed for the purpose of security, and
3. the battery gap, which is a consequence of the energy consumed in performing security-related functions.
We will provide an overview of our research in this area, covering both embedded system architectures that address these gaps, and methodologies that assist in their design. We will use mobile appliances (mobile phones, PDAs) to illustrate secure embedded system design challenges, and describe MOSES, a security platform that we have developed and deployed in NEC's next-generation mobile phones.
Several technologies have been developed to address information security (cryptography, secure communication protocols, anti-virus tools, firewalls, intrusion detection, and so on), which can be adapted to embedded systems. These technologies can be referred to as "functional" security meas-ures, since they usually specify functions that must be added to the target system without any consid-eration of how they are embodied in hardware or software.
However, they are hardly sufficient to ensure the security of embedded systems in practice. Most real security attacks do not directly take on the theoretical strength of cryptographic algorithms; instead, they target weaknesses in a system's "implementation". Moreover, embedded-system designers have to cope with security as yet another requirement, in addition to performance, power, cost, etc.
We will present an introduction to embedded system security challenges, and argue that ef-fective security solutions can be realized only if they are built-in at various stages of the design process (architecture, HW design, and SW development). The objectives of secure embedded system design will be defined from the designer's perspective as addressing various "gaps" such as
1. the assurance gap, which refers to the gap between functional security measures and truly secure implementations,
2. the security processing gap, which arises due to the processing requirements of the additional computations that must be performed for the purpose of security, and
3. the battery gap, which is a consequence of the energy consumed in performing security-related functions.
We will provide an overview of our research in this area, covering both embedded system architectures that address these gaps, and methodologies that assist in their design. We will use mobile appliances (mobile phones, PDAs) to illustrate secure embedded system design challenges, and describe MOSES, a security platform that we have developed and deployed in NEC's next-generation mobile phones.
No comments:
Post a Comment