• No results found

// P a t c h for the ’ qemu - 2 . 1 2 . 0 - rc0 / t a r g e t / arm / h e l p e r . c ’ f i l e

... u i n t 6 4 _ t f a k e _ c o u n t e r = 0; ... v o i d p m c c n t r _ s y n c ( C P U A R M S t a t e * env ) { u i n t 6 4 _ t t e m p _ t i c k s ; t e m p _ t i c k s = m u l d i v 6 4 ( q e m u _ c l o c k _ g e t _ n s ( Q E M U _ C L O C K _ V I R T U A L ) , A R M _ C P U _ F R E Q , N A N O S E C O N D S _ P E R _ S E C O N D ) ; if ( env - > c p 1 5 . c 9 _ p m c r & P M C R D ) { /* I n c r e m e n t o n c e e v e r y 64 p r o c e s s o r c l o c k c y c l e s */ t e m p _ t i c k s /= 64; } if ( a r m _ c c n t _ e n a b l e d ( env ) ) { env - > c p 1 5 . c 1 5 _ c c n t = t e m p _ t i c k s - env - > c p 1 5 . c 1 5 _ c c n t ; } f a k e _ c o u n t e r = 0; }

s t a t i c u i n t 6 4 _ t p m c c n t r _ r e a d ( C P U A R M S t a t e * env , c o n s t A R M C P R e g I n f o * ri ) { u i n t 6 4 _ t t o t a l _ t i c k s ; if (! a r m _ c c n t _ e n a b l e d ( env ) ) { /* C o u n t e r is d i s a b l e d , do not c h a n g e v a l u e */ r e t u r n env - > c p 1 5 . c 1 5 _ c c n t ; } t o t a l _ t i c k s = m u l d i v 6 4 ( q e m u _ c l o c k _ g e t _ n s ( Q E M U _ C L O C K _ V I R T U A L ) , A R M _ C P U _ F R E Q , N A N O S E C O N D S _ P E R _ S E C O N D ) ; if ( env - > c p 1 5 . c 9 _ p m c r & P M C R D ) { /* I n c r e m e n t o n c e e v e r y 64 p r o c e s s o r c l o c k c y c l e s */ t o t a l _ t i c k s /= 64; } if ( f a k e _ c o u n t e r == 0 || f a k e _ c o u n t e r == 3) t o t a l _ t i c k s = 2; e l s e if ( f a k e _ c o u n t e r == 1) t o t a l _ t i c k s = 3; e l s e if ( f a k e _ c o u n t e r == 2) t o t a l _ t i c k s = 0; e l s e if ( f a k e _ c o u n t e r == 4) t o t a l _ t i c k s = 4; f a k e _ c o u n t e r ++; r e t u r n t o t a l _ t i c k s - env - > c p 1 5 . c 1 5 _ c c n t ; } ...

Appendices

B.1

Performance Results from MEDA, k -NN and RF

Table B.1: MEDA performance results on recognising REMs and VESs for 1000 pings (dataset of REMs from Chapter 3, Section 3.4.2).

Precision Recall F1-score Accuracy ODP REM 1.0000 0.4455 0.6164

0.7921 78.03% VES 0.7504 1.0000 0.8574

Average 0.8752 0.7228 0.7369

Table B.2: MEDA performance results on recognising REMs and VESs for 25 pings.

Precision Recall F1-score Accuracy ODP REM 0.5340 1.0000 0.6962

0.6728 71.78% VES 1.0000 0.4764 0.6454

Average 0.7670 0.7382 0.6708

Table B.3: ERT+k -NN performance results on recognising REMs and VESs for 25 pings.

Precision Recall F1-score Accuracy ODP REM 0.9878 0.9810 0.9844

0.9846 98.46% VES 0.9815 0.9881 0.9848

Average 0.9847 0.9846 0.9846

Table B.4: L1-FS+k -NN performance results on recognising REMs and VESs for 25 pings.

Precision Recall F1-score Accuracy ODP REM 0.9842 0.9814 0.9828

0.9830 98.30% VES 0.9818 0.9845 0.9832

Average 0.9830 0.9830 0.9830

Table B.5: ERT+RF performance results on recognising REMs and VESs for 25 pings.

Precision Recall F1-score Accuracy ODP REM 0.9968 0.9952 0.9960

0.9960 99.60% VES 0.9953 0.9968 0.9960

Average 0.9960 0.9960 0.9960

Table B.6: L1-FS+RF performance results on recognising REMs and VESs for 25 pings.

Precision Recall F1-score Accuracy ODP REM 0.9968 0.9952 0.9960

0.9960 99.60% VES 0.9953 0.9968 0.9960

Average 0.9960 0.9960 0.9960

Table B.7: MEDA performance results on recognising REMs and VESs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 0.6083 1.0000 0.7564

0.7585 78.53% VES 1.0000 0.6136 0.7605

Average 0.8041 0.8068 0.7585

Table B.8: ERT+k -NN performance results on recognising REMs and VESs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 0.9900 0.9964 0.9932

0.9932 99.32% VES 0.9964 0.9901 0.9932

Table B.9: L1-FS+k -NN performance results on recognising REMs and VESs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 0.9955 0.9899 0.9927

0.9928 99.28% VES 0.9901 0.9956 0.9929

Average 0.9928 0.9928 0.9928

Table B.10: ERT+RF performance results on recognising REMs and VESs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 0.9996 0.9976 0.9986

0.9986 99.86% VES 0.9976 0.9996 0.9986

Average 0.9986 0.9986 0.9986

Table B.11: L1-FS+RF performance results on recognising REMs and VESs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 0.9992 0.9976 0.9984

0.9984 99.84% VES 0.9976 0.9992 0.9984

Average 0.9984 0.9984 0.9984

Table B.12: MEDA performance results on recognising unknown REMs and VESs for 25 pings.

Precision Recall F1-score Accuracy ODP REM 0.0000 0.5000 0.0000

0.5000 57.14% VES 1.0000 1.0000 1.0000

Average 0.5000 0.7500 0.5000

Table B.13: ERT+k -NN performance results on recognising unknown REMs and VESs for 25 pings.

Precision Recall F1-score Accuracy ODP REM 1.0000 0.8075 0.8821

0.9005 94.00% VES 1.0000 0.9935 0.9967

Table B.14: L1-FS+k -NN performance results on recognising unknown REMs and VESs for 25 pings.

Precision Recall F1-score Accuracy ODP REM 1.0000 0.8960 0.9420

0.9360 96.26% VES 1.0000 0.9760 0.9879

Average 1.0000 0.9360 0.9649

Table B.15: ERT+RF performance results on recognising unknown REMs and VESs for 25 pings.

Precision Recall F1-score Accuracy ODP REM 1.0000 0.9585 0.9786

0.9750 98.56% VES 1.0000 0.9915 0.9957

Average 1.0000 0.9750 0.9872

Table B.16: L1-FS+RF performance results on recognising unknown REMs and VESs for 25 pings.

Precision Recall F1-score Accuracy ODP REM 1.0000 0.4008 0.5723

0.9850 99.14% VES 1.0000 0.9935 0.9967

Average 1.0000 0.6972 0.7845

Table B.17: MEDA performance results on recognising unknown REMs and VESs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 0.7750 0.5005 0.2760

0.5003 72.17% VES 1.0000 1.0000 1.0000

Average 0.8875 0.7503 0.6380

Table B.18: ERT+k -NN performance results on recognising unknown REMs and VESs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 1.0000 0.9775 0.9885

0.9783 98.75% VES 1.0000 0.9790 0.9893

Table B.19: L1-FS+k -NN performance results on recognising unknown REMs and VESs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 1.0000 0.9635 0.9811

0.9793 98.80% VES 1.0000 0.9950 0.9975

Average 1.0000 0.9793 0.9893

Table B.20: ERT+RF performance results on recognising unknown REMs and VESs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 1.0000 0.9995 0.9997

0.9990 99.94% VES 1.0000 0.9985 0.9992

Average 1.0000 0.9990 0.9995

Table B.21: L1-FS+RF performance results on recognising unknown REMs and VESs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 1.0000 0.9410 0.9687

0.9700 98.26% VES 1.0000 0.9990 0.9995

Average 1.0000 0.9700 0.9841

Table B.22: k -NN performance results on recognising REMs and FTAs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 0.0000 0.0600 0.0011

0.9410 95.44% FTA 1.0000 0.9400 0.9700

Average 1.0000 0.0600 0.1100

Table B.23: RF performance results on recognising REMs and FTAs for 200 pings.

Precision Recall F1-score Accuracy ODP REM 0.0000 0.6500 0.2100

0.3480 53.97% FTA 1.0000 0.3500 0.5200

B.2

Performance Results from k -NN and RF without the