Power Analysis Attacks -
Revealing the Secrets of Smartcards
DPAbook Errata  
 
About the DPA Book
- Abstract
- Table of Contents
- Release Date
- Authors

Errata
- Online Errata

Online Material
- Matlab Scripts
- Matlab Workspaces

Links
- Universities
- Industry
- Conferences



Errata

This is a list of typos and other errors that have been found in the book. It is updated continuously.

If you find any typo or error that is not already in the list, please report it to errata@dpabook.org.

Page Exact Location Text in the Book Text Should Be Reported by
xviii Line 13 Φ(x)     Probability density function of ... Φ(x)     Cumulative distribution function of ... Thomas Popp
65 Paragraph 3, Line 6 Hence, Var(Pdata) and Var(Pel.noise) ... Hence, Var(Pop) and Var(Pdata) ... Thomas Popp
91 Paragraph 2, Line 3 ... if only S is known ... the variable T = (X_bar - μ) · sqrt(n) / S is ... ... if only s is known ... the variable T = (X_bar - μ) · sqrt(n) / s is ... Thomas Popp
92 Second to last paragraph, Lines 6-8 Hence, we want to have that p((X_bar - μ) · σ / sqrt(n) < -μ · σ / sqrt(n)) = 1 - α. This allows deriving the number of traces n: p((X_bar - μ) · σ / sqrt(n) < -μ · σ / sqrt(n)) = Φ(-μ · σ / sqrt(n)) = 1 - α. Hence, we want to have that p((X_bar - μ) · sqrt(n) / σ < -μ · sqrt(n) / σ) = 1 - α. This allows deriving the number of traces n: p((X_bar - μ) · sqrt(n) / σ < -μ · sqrt(n) / σ) = Φ(-μ · sqrt(n) / σ) = 1 - α. Stefan Tillich
93 Paragraph 3, Line 2 ... with precision c = 0.01 is 176 306. ... with precision c = 0.01mV is 176 306. Thomas Popp
93 Paragraph 3, Line 3, Word 6 N(1.86,1.63) N(111.86,1.63) Stefan Tillich
93 Paragraph 3, Line 3-4 ... with confidence 0.99 is n = 4. ... with confidence 0.99 is n = 1 (the actual value of n is 0.00115, however, it can only be a positive integer). This means that already one trace is more than enough to tell with very high confidence that the mean of the distribution is different from zero, i.e. the hypothesis μ_0 = 0 is already part of the critical region in this case with very high probability and is thus rejected. Yu Yu
95 Paragraph 1, Line 2 (if not, we just switch X and Y). (if not, we just switch X_bar and Y_bar). Stefan Tillich
95 Paragraph 1, Lines 3-5 We know that p(Z < 0) equals Φ(-(μ_X - μ_Y) · sqrt(n) / (2 / sqrt(σ))) and therefore we have Φ(-(μ_X - μ_Y) · sqrt(n) / (2 / sqrt(σ))) = 1 - α. We know that p(Z < 0) equals Φ(-(μ_X - μ_Y) · sqrt(n) / (2 · σ)) and therefore we have Φ(-(μ_X - μ_Y) · sqrt(n) / (2 · σ)) = 1 - α. Stefan Tillich
95 Lines 2/3 of "Important Box" ... X ~ N(μ_X, σ / sqrt(n/2)) and Y ~ N(μ_Y, σ / sqrt(n/2)) ... ... X_bar ~ N(μ_X, σ / sqrt(n/2)) and Y_bar ~ N(μ_Y, σ / sqrt(n/2)) ... Stefan Tillich
95 Paragraph 4, Line 10 ... confidence interval with c = 1, about n = 4823 traces would be ... ... confidence interval with c = 1mV, about n = 2412 traces would be ... Thomas Popp
95 Paragraph 5, Line 4 ... we need n = 112 traces ... ... we need n = 56 traces ... Thomas Popp
148 Line 1 of "Important Box" For |ρ_{ck,ct} <= 0.2| and ... For |ρ_{ck,ct}| <= 0.2 and ... Thomas Popp
168 "Important Box" at the bottom The power consumption a cryptographic device ... The power consumption of a cryptographic device ... Thomas Popp
233 Paragraph 1, Lines 4-5 ... (notice that this a random result) ... ... (notice that this is a random result) ... Thomas Popp
241 Paragraph 2, Line 5 ... performed by the two SR AND cells, the two SR OR cells, and the ... ... performed by the SR AND cell, the SR OR cell, and the ... Thomas Popp
241 Paragraph 2, Line 8 It calculates as output d_{m(t+1)} = d XOR m(t+1) and its inverse. It calculates as output d_{m(t+1)} = d XOR m(t+1). Thomas Popp
249 Last paragraph, Line 6 ... to recompute a the masked table ... ... to recompute the masked table ... Thomas Popp
252 Paragraph 3, Line 2, Word 7 ρ(H, tilde{T}) = ρ(HW(V XOR W), tilde{T}) ρ(H, tilde{T}) = ρ(HW(U XOR V), tilde{T}) Stefan Tillich
252 Equation 10.2, at the end , HW(v_m)) , HW(v_m))) Stefan Tillich
258 The three equations comb(u, v) = - 89.95 · sin(HW(u XOR v)^3) - - 7.82 · sin(HW(u XOR v)^2) + 67.66

pre(HW(u_m), HW(v_m)) = sin(HW(u_m) - HW(v_m))^2

ρ(comb(u,v), pre(HW(u_m),HW(v_m))) = 0.83
comb(u, v) = - 89.95 · sin(HW(u XOR v))^3 - 7.82 · sin(HW(u XOR v))^2 + 67.66 · sin(HW(u XOR v))

pre(HW(u_m), HW(v_m)) = sin((HW(u_m) - HW(v_m))^2)

ρ(comb(u,v), pre(HW(u_m),HW(v_m))) = 0.84
Stefan Tillich, Thomas Popp
258 Paragraph 4, Line 6, Word 3 u = HW((d_i XOR k_j) u = d_i XOR k_j Stefan Tillich
   
last update: 25.01.2010 Sponsored by Graz University of Technology