Software Failure Modes
Unsafe Control Action
Software has different modes of failure. The conditional probability of software failure given the existence of certain defect types, has been discussed and developed in previous work by Idaho National Laboratory (see [1]). Data from [5] and [6] were assessed in terms of their impact on the usability, functionality, and performance from an end-user perspective, and were assigned a unsafe control action (UCA) or unsafe information flow (UIF) category (equivalently, a mode of failure). The assessment result provided conditional relationships, shown in Table 4. More specifically, Table 4, which provides a global perspective on the relationship between defect types and UCA/UIF failure modes [1]. As further classifications are performed, these distributions can be continually improved. These correlations are used by Eq. 13, where is the total number of defect types considered by BAHAMAS, and is the UCA/UIF failure mode being evaluated. Table 4 provides the conditional probabilities used for:
(13)P(software\ failure\ of\ mode\ x) = \sum_{i=1}^{I} P(UC_x|Type_i)P(Type_i)
An inherent assumption in this work is that software failure results from active defects. An implication of Eq. 13 is that the defect activation probability equals one, and that the activation of defects is mutually exclusive (i.e., multiple defects are not activated simultaneously). In other words, the probability of software failure is a summation of the contribution of each defect type. Future work may investigate refinements to this activation probability.
Defect Class |
UCA/UIF |
UCA/UIF |
UCA/UIF |
UCA/UIF |
|---|---|---|---|---|
A |
B |
C |
D |
|
Algorithm |
0.217±0.051 |
0.525±0.068 |
0.124±0.044 |
0.134±0.019 |
Assignment |
0.288±0.152 |
0.667±0.130 |
0.045±0.072 |
N/A |
Checking |
0.219±0.074 |
0.539±0.115 |
0.102±0.072 |
0.141±0.129 |
Documentation |
0.250±0.250 |
0.250±0.250 |
0.250±0.250 |
0.250±0.250 |
Function |
0.250±0.154 |
0.518±0.064 |
0.157±0.092 |
0.074±0.216 |
Interface |
0.262±0.065 |
0.579±0.086 |
0.093±0.095 |
0.065±0.039 |
Relationship |
0.250±0.250 |
0.250±0.250 |
0.250±0.250 |
0.250±0.250 |
Timing |
0.095±0.334 |
0.190±0.289 |
0.524±0.423 |
0.190±0.289 |
Note: This table was based on [1]. These distributions may change as more data are collected and data collection methods improve.