US 11,720,462 B2
Physical execution monitor
Patrick Thomas Michael Klapper, Schwalbach a. Ts. (DE); Marc Sebastian Patric Stöttinger, Schwalbach a. Ts. (DE); and Miguel Hernandez, Schwalbach a. Ts. (DE)
Assigned to CONTINENTAL TEVES AG & CO. OHG, Frankfurt am Main (DE)
Filed by Continental Teves AG & Co. OHG, Frankfurt am Main (DE)
Filed on Jul. 29, 2020, as Appl. No. 16/941,723.
Claims priority of application No. 19189184 (EP), filed on Jul. 30, 2019.
Prior Publication US 2021/0034489 A1, Feb. 4, 2021
Int. Cl. G06F 11/00 (2006.01); G06F 11/30 (2006.01); G06F 17/16 (2006.01); G06F 18/24 (2023.01); G06F 21/57 (2013.01)
CPC G06F 11/3024 (2013.01) [G06F 17/16 (2013.01); G06F 18/24 (2023.01); G06F 21/57 (2013.01)] 13 Claims
OG exemplary drawing
 
1. A computer-implemented method of monitoring execution of computer program instructions in a microprocessor of a digital system, comprising:
a training phase comprising:
determining input data reference values for physical properties for a training phase;
generating fingerprints of measured properties; and
creating templates for identifying irregularities in operations based on the fingerprints;
the monitoring comprising:
receiving a plurality of data items representing real-time measurements of side-channel information emanating from and correlated to the execution of computer program instructions in the monitored digital system, each one of the plurality of data items being sampled in a timed relationship or correlated to each other and forming a value of a corresponding dimension of a side-channel information vector;
receiving, for two or more of the corresponding dimensions of the side-channel information vector, classifiers that assign a received side-channel vector to one of two classes;
classifying the received side-channel information vector in accordance with respective received classifiers, wherein an orthogonal distance of the side-channel information vector from a corresponding classifier indicates an associated probability or confidence value of the respective classification;
generating a combined probability or confidence value for the side-channel information vector from individual classifications and associated probability or confidence values; and
outputting a signal if a combined probability or confidence value of a side-channel information vector indicates affiliation to a selected one of the two classes with a predetermined probability.