Sig­nal pro­cess­ing and AI-​based data eval­u­a­tion

In order to re­al­ize an ef­fi­cient data eval­u­a­tion in terms of Green ICT and under real-​time con­di­tions we in­ves­ti­gate sig­nal pro­cess­ing and AI meth­ods which can be used for resource-​limited sys­tems. Es­pe­cially for AI meth­ods the em­pha­sis is on the re­al­iza­tion of hard­ware ac­cel­er­a­tors, e.g. for the de­tec­tion of at­tacks on the basis of proces­sor data or for the de­tec­tion of freez­ing of gait and the trig­ger­ing of counter mea­sures within less than 10 mil­lisec­onds for Parkin­son pa­tients. These ap­proaches are also used to eval­u­ate the con­di­tion of the sys­tems in order to con­trol their re­silience by ini­ti­at­ing ap­pro­pri­ate mea­sures such as re­lo­ca­tion of ac­tiv­i­ties to neigh­bor­ing sys­tems, restarts for rather tran­sient faults, etc.