As computers continue to pervade all walks of our lives, dependability (an umbrella term for reliability, timeliness and security) is becoming the most wanted property which ultimately will surpass in significance insatiable appetite for computer performance.
Starting with my own perspective on modeling, I will introduce dependability measures and then focus on dependability models including fault, system and service models. Next, analytical, simulation and experimental approaches will be discussed. The methods range from Reliability Block Diagrams and Fault Trees to Petri Nets and/or Markov Chains. I will explain how one approaches key issues in the dependability modeling and then present a brief survey of modeling tools.
The lecture ends with our approach to dependability modeling of a complex software/hardware communication platform and a list of current challenges, including models composability.