Reliability metrics

Numeric solutions for PFD, PFS, MTTF and other reliability metrics can be obtained from this matrix using a spreadsheet. 

Recovery rates for the router crash scenario are given in Figure 5 in terms of the reliability metric. Kademlia provides success rates above 75% with 16 peers or more. Chord ranges between 65% and 95% with 16 peers or more. Recovery failures occur due to two different causes (i) The P2P network communication is partially disturbed due to the router crash and therefore unable to satisfy all requests or (ii) P2P messages get lost due to the packet error rate. Fluctuations in the success rates occur due to PeSCADA s replication and routing scheme In case the RTUs that are affected by the router crash are responsible for the specific address space range of the requested datum, the MTU cannot retrieve the... 

Predictive Models The Rome Air Development Center Reliability Metric... 

Harrington and Isenhour developed a quantitative reliability metric (QRM) for determining the reliability of library searching methods with infrared spectra. They report applying the metric for evaluating the reliability of library searches for unknown target spectra and using the measure for de-... 

Establishing the Correlation between Complexity and a Reliability Metric for Software Digital I C-Systems... 

The purpose of the approach is to create a link between available system information and a reliability metric. The aim is not to find the true reliability of a system but rather use the reliability metric as a mean or a tool for adjusting the analysis work. The methods for identifying and collecting system information was described in Chapter 2 and they are performed in the three first work packages of the project. For the creation of a coimection between the input and the reliability metric there are several viable approaches. 

The available input metrics (as mentioned in Chapter 2.4) were combined in three branches in the BBN, each using different input metrics in their top nodes and eventually, through a series of intermediate nodes, end in one leaf node describing the reliability metric. The experts decided the combination of inputs into the intermediate nodes. 

LogicaLComplexity-Variability. The first branch is from the last intermediate node connected to the node representing the reliability metric, namely Reliabil-ity.Metric. 

The second branch represents the connectivity within the LD. The first input node, Number-FDLFDO, represents the number of connected LDs to the input and number of connected LDs to the output of the LD. The second input node, Number-Input-Output, is the total inputs and outputs to the LD. The two input nodes were joined by the experts in the intermediate node Connectivity which represents the number of connected inputs and outputs to the LD in total, and the number of other LDs connected. The intermediate node Connectivity was connected to the reliability metric node. 

A case was presented where inputs, based on metrics available fi om a set of FBs and LDs specifying a digital I C-System, were combined by experts in a BBN to express a relationship between complexity of a logic diagram and a reliability metric for that diagram. In the project, data from system specification have been extracted and complexity calculation for the metrics has been performed. A meeting with experts was facilitated for the creation of a Bayesian Belief Net connecting complexity with a reliability metric. 

To circumvent the coverage problem some approaches advocate to continuously monitor the used stack during normal operation of the system for a given period in time, aiming at a reliability metrics based on the time spent for measurements. However, in contrast to hardware metrics, the results are inconclusive since there is no indication how often a specific execution path has been exercised during the observation period, or whether it has been exercised at all. In consequence for software-based systems no statistical failure rates are available which are comparable to those used for hardware... 

DFT Analysis. DFTCalc can compute a number of different reliability metrics, namely all metrics that can be expressed as reachability properties in the logic CSL. This includes properties such as (1) Timed-Reliability the probability that the system fails until a given time point T or in a given interval [T, T ] (2) Mean time to failure the expected time to a system failure (3) Reliability the probability that the system fails in the long-run. In case of non-determinism, we calculate the minimum and maximum values for the above metrics. Each of these properties can either be evaluated from the initial state (i.e. the system is fully functional), or by setting evidence (i.e certain components have failed already). 

---