The pervasiveness of modern day embedded systems has led to the storing of huge amount of sensitive information in them. These embedded devices have to often operate under insecure environments and hence are susceptible to software and physical attacks. Hence security becomes a prime concern in embedded systems. Although a lot of hardware cryptographic techniques have been proposed to provide high levels of security, they are hampered by the trade-offs created by the energy constraints in embedded systems. In this paper, we propose an Energy Efficient Memory Integrity Verification Mechanism that can adaptively tune a Memory Integrity Verification Module(MIV) to a Sensor Module(SM). This drastically reduces the energy overheads imposed on an embedded system as compared to the conventional security mechanisms. The simulation results help us conclude that the average energy saved in our mechanism ranges from 88% to 99%. This is much higher as compared to the results achieved in baseline simulations with traditional memory integrity verification techniques.
Computer security aims at protecting confidentiality, integrity, and availability of sensitive information that are processed, used, or stored by computing systems. Computer scientists working in the field of computer security have successfully designed and developed software and hardware mechanisms to provide security in modern day computing devices. As compared to hardware security mechanisms, software-only security mechanisms are easy to implement and patch. But software-only security mechanisms cannot ensure protection against hardware-based attacks, thus rendering them vulnerable to such attacks. Hardware mechanism such as secure architectures aim to root the trust of the security solution in the hardware architecture. These security architectures typically deploy security mechanisms like encryption/decryption to protect confidentiality and hashing to protect data integrity. Though the security provided by hardware secure architectures is reliably high, they require modifications to the processor micro-architecture. Any changes to the micro-architecture is an extremely costly and time consuming process. Also, testing these hardware secure architectures is difficult as it requires testing the complete system including hardware, software and applications. Recently, virtualization has emerged to be an efficient and cost effective technology that allows emulating hardware mechanisms. It also enables emulating new hardware features in a virtualized environment. This makes the task of testing security architectures efficient and easy. In this paper, we use a virtualization software to build a Virtualization Based Secure Execution and Testing Framework for testing hardware secure architectures. Our framework provides a mechanism to plug-in secure architectures and monitor or test the system behavior by performing attacks on it.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.