. Author manuscript; available in PMC: 2021 May 22.

Published in final edited form as: J Rehabil Res Dev. 2011;48(6):xxi–xxxviii. doi: 10.1682/jrrd.2011.03.0055

Table 1.

Worse-case, average, and best-case delay equations with delay ranges.

Classifier Type	Worst-Case Delay	Average Delay	Best-Case Delay	Difference Between Best & Worst Cases
No Overlap, No Majority Voting	$D = \frac{3}{2} T_{a} + τ$	$D = T_{a} + τ$	$D = \frac{1}{2} T_{a} + τ$	T_a
No Overlap, with Majority Voting	$D = (\frac{n}{2} + 1) T_{a} + τ$	$D = (\frac{n + 1}{2}) T_{a} + τ$	$D = (\frac{n}{2}) T_{a} + τ$	T_a
Overlap, No Majority Voting	$D = \frac{1}{2} T_{a} + T_{new} + τ$	$D = \frac{1}{2} T_{a} + \frac{1}{2} T_{new} + τ$	$D = \frac{1}{2} T_{a} + τ$	T_new
Overlap, With Majority Voting	$D = \frac{1}{2} T_{a} + (\frac{n + 1}{2}) T_{new} + τ$	$D = \frac{1}{2} T_{a} + (\frac{n}{2}) T_{new} + τ$	$D = \frac{1}{2} T_{a} + (\frac{n - 1}{2}) T_{new} + τ$	T_new

Note: Equations to estimate worst-case, average, and best-case controller delay as well as difference between best- and worst-case controller delays. General rule of thumb is that T_new is approximately an order of magnitude less than T_a, meaning that users should experience more consistent delay with overlapped windows.

D = maximum delay between users intended movement and controller producing correct output class, n = number of majority votes, T_a = analysis window length, τ = processing time, T_new = amount of window overlap.