Re­silient Sys­tems

    The term re­silience is used in many sci­en­tific dis­ci­plines rang­ing from ma­te­ri­als to psy­chol­ogy.

    We are in­ter­ested in the re­silience of tech­ni­cal sys­tems, fo­cus­ing on cyber phys­i­cal sys­tems of sys­tems (CPSoS). We are re­search­ing op­tions to mea­sure re­silience, de­sign and test re­silient CPSoS. We de­fine CPS(oS) as re­silient, if they have the abil­ity to react to spec­i­fied and un­spec­i­fied dis­tur­bances in a way that pre­serves their func­tion and en­ables them to react quickly. This re­ac­tion in­cludes the early de­tec­tion, min­i­miza­tion, pre­dic­tion or even avoid­ance of dis­tur­bances. The IHP has the fol­low­ing com­pe­ten­cies in the field of re­silient sys­tems:

    Com­pe­ten­cies

    • En­ergy Ef­fi­cient Im­ple­men­ta­tions

      En­ergy Ef­fi­cient Im­ple­men­ta­tions

      An im­por­tant strat­egy to reach re­silience is re­dun­dency. In the ap­pli­ca­tion fields con­sid­ered here such as e-​health, au­toma­tion net­works, etc., em­bed­ded sys­tems are gen­er­ally used.

      more In­for­ma­tion

    • Tam­per Re­sis­tant Crypto Im­ple­men­ta­tions

      Tam­per Re­sis­tant Crypto Im­ple­men­ta­tions

      The pro­tec­tion goals con­fi­den­tial­ity, data in­tegrity and au­then­tic­ity can be re­al­ized with the help of cryp­to­graphic al­go­rithms. To achieve this, it is es­sen­tial to keep the used keys se­cret.

      more In­for­ma­tion

    • At­tack De­tec­tion

      At­tack De­tec­tion

      The threat to IT sys­tems from at­tacks is con­tin­u­ously in­creas­ing. The de­tec­tion of at­tacks is very chal­leng­ing es­pe­cially for em­bed­ded sys­tems as they have very lim­ited re­sources.

      more In­for­ma­tion

    • Rad­hard RF and Dig­i­tal Cir­cuits for Re­li­a­bil­ity Crit­i­cal Ap­pli­ca­tions

      Rad­hard RF and Dig­i­tal Cir­cuits for Re­li­a­bil­ity Crit­i­cal Ap­pli­ca­tions

      Re­search ac­tiv­i­ties for re­li­able dig­i­tal and RF cir­cuits for space ap­pli­ca­tion have long tra­di­tion. The major focus of this in­ves­ti­ga­tion is on ra­di­a­tion hard by de­sign (RHBD) method­ol­ogy for dig­i­tal and RF cir­cuits.

      more In­for­ma­tion

    • Adap­tive Fault Tol­er­ant Multi-​Processors

      Adap­tive Fault Tol­er­ant Multi-​Processors

      Pro­cess­ing is one of the crit­i­cal op­er­a­tions in em­bed­ded sys­tems with re­li­a­bil­ity rel­e­vant re­quire­ments. IHP is in­ves­ti­gat­ing single-​​ and multi-​​processing sys­tems that pro­vide re­li­able op­er­a­tion for dif­fer­ent ap­pli­ca­tions, in­clud­ing space, au­to­mated dri­ving, or in­dus­try au­toma­tion.

      more In­for­ma­tion

    • Re­li­able Com­put­ing Ac­cel­er­a­tion for AI and DSP ap­pli­ca­tions

      Re­li­able Com­put­ing Ac­cel­er­a­tion for AI and DSP ap­pli­ca­tions

      Today AI ap­pli­ca­tions are dri­ving tech­nol­ogy de­vel­op­ment. In order to ful­fill per­for­mance and en­ergy re­quire­ments, in­no­v­a­tive so­lu­tions are needed.

      more In­for­ma­tion

    Re­silience is more than se­cu­rity and re­dun­dancy. Cog­ni­tion makes the dif­fer­ence.

    Se­cu­rity is THE pre­req­ui­site needed to en­sure any fea­ture of a cer­tain sys­tem. With­out se­cu­rity an at­tacker can take con­trol of the sys­tem and in con­se­quence all its fea­tures are sub­ject to the in­ten­tion of the at­tacker. 

    Re­dun­dancy is a well-​known op­tion to in­crease the re­li­a­bil­ity of al­most any sys­tem. With­out a cer­tain level of re­dun­dancy a sys­tem will fail in the mo­ment at which one of its es­sen­tial com­po­nents breaks. So at the least these com­po­nents re­quire re­dun­dancy. But the af­ford­able re­dun­dancy is lim­ited, e.g. 2 en­gines for cer­tain types of air­planes. But what hap­pens if the re­dun­dancy is lost, e.g. as in the case when an air­plane ex­pe­ri­ences a dou­ble bird strike shortly after take­off.  

    Cop­ing with the Un­ex­pected 

    What is going to help in an un­ex­pected sit­u­a­tion such as the dou­ble bird strike?

    The an­swer is cog­ni­tion, i.e. find­ing a so­lu­tion to a for­merly un­ex­pe­ri­enced prob­lem. As we are work­ing on CPSoS the chal­lenge is not only to come up with a sys­tem that  re­acts prop­erly to such a chal­lenge, but to re­al­ize it with the very lim­ited re­sources of em­bed­ded sys­tems.

    So metaphor­i­cally speak­ing, the prob­lem we are try­ing  to solve is squeez­ing Cap­tain Sul­len­berger (the pilot who landed the air­plane on the Hud­son river) into mil­lions of em­bed­ded sys­tems.

    Prof. Peter Langendörfer

    De­part­ment Head

    IHP
    Im Tech­nolo­giepark 25
    15236 Frank­furt (Oder)
    Ger­many

    Sec­re­tary:
    Ste­fanie Fro­nia
    Phone: +49 335 5625 720
    Fax: +49 335 5625 671
    Send e-​mail »

    The website is designed for modern browsers. Please use a current browser.