TY - GEN
T1 - Secure erasure and code update in legacy sensors
AU - Karame, Ghassan O.
AU - Li, Wenting
PY - 2015
Y1 - 2015
N2 - Sensors require frequent over-the-air reprogramming to patch software errors, replace code, change sensor configuration, etc. Given their limited computational capability, one of the few workable techniques to secure code update in legacy sensors would be to execute Proofs of Secure Erasure (PoSE) which ensure that the sensor’s memory is purged before sending the updated code. By doing so, the updated code can be loaded onto the sensor with the assurance that no other malicious code is being stored. Although current PoSE proposals rely on relatively simple cryptographic constructs, they still result in considerable energy and time overhead in existing legacy sensors. In this paper, we propose a secure code update protocol which considerably reduces the overhead of existing proposals. Our proposal naturally combines PoSE with All or Nothing Transforms (AONT); we analyze the security of our scheme and evaluate its performance by means of implementation on MicaZ motes. Our prototype implementation only consumes 371 bytes of RAM in TinyOS2, and improves the time and energy overhead of existing proposals based on PoSE by almost 75 %.
AB - Sensors require frequent over-the-air reprogramming to patch software errors, replace code, change sensor configuration, etc. Given their limited computational capability, one of the few workable techniques to secure code update in legacy sensors would be to execute Proofs of Secure Erasure (PoSE) which ensure that the sensor’s memory is purged before sending the updated code. By doing so, the updated code can be loaded onto the sensor with the assurance that no other malicious code is being stored. Although current PoSE proposals rely on relatively simple cryptographic constructs, they still result in considerable energy and time overhead in existing legacy sensors. In this paper, we propose a secure code update protocol which considerably reduces the overhead of existing proposals. Our proposal naturally combines PoSE with All or Nothing Transforms (AONT); we analyze the security of our scheme and evaluate its performance by means of implementation on MicaZ motes. Our prototype implementation only consumes 371 bytes of RAM in TinyOS2, and improves the time and energy overhead of existing proposals based on PoSE by almost 75 %.
KW - All or nothing transformations
KW - Proofs of secure erasure
KW - Secure code update
UR - http://www.scopus.com/inward/record.url?scp=84944574137&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-22846-4_17
DO - 10.1007/978-3-319-22846-4_17
M3 - Conference contribution
AN - SCOPUS:84944574137
SN - 9783319228457
VL - 9229
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 283
EP - 299
BT - Trust and Trustworthy Computing - 8th International Conference, TRUST 2015, Proceedings
PB - Springer Verlag
T2 - 8th International Conference on Trust and Trustworthy Computing, TRUST 2015
Y2 - 24 August 2015 through 26 August 2015
ER -