## Errata

__Paper title:__ Impact of Integrity Attacks on Real-Time Pricing in Smart Grids

__Venue:__ ACM CCS 2013
- Page 8, right column, 6th line of Section 5.3:

*G*_{w}(z) should be *G*_{w}(z) = z^{-τ} ρ ẇ(λ_{o}) + (1-ρ) ẇ(λ_{o})

The above typo does not affect any results in the paper.

*last updated @ Apr 27, 2014*

__Paper title:__ Impact of Integrity Attacks on Real-Time Pricing in Smart Grids

__Venue:__ ADSC Tech report, 2013
- Page 9, left column, 4th line of Section V-C:

*W(z)* should be *W(z) = z*^{-τ} ρ ẇ(λ_{o}) + (1-ρ) ẇ(λ_{o})
- Page 13, right column, 6th line of Appendix C:

*T*_{c}(z) should be *2ηz / (z - 1 + 2η)*

The above typos do not affect any results in the paper.

*last updated @ Apr 27, 2014*

A discussion in the work on *adaptive calibration* was inaccurate. This discussion is
Particularly, for our problem, the system converges faster for larger *a* at the price of worse system stability. Therefore, the best setting for *a* is a value close to the upper bound *σ*^{2}/S.

This discussion can be found in the following papers:
The correct statement should be:
Particularly, for our problem, when *a=σ*^{2}/(2S), the system converges fastest.

The inaccurate discussion in the papers do not affect any results, since *β* is set to be 0.5 in the simulations and experiments, which coincides the true optimal setting for *a* in the corrected statement.

*last updated @ Apr 28, 2014*