Reliability and performance based measures for computer networks and distributed systems

A novel reliability measure which incorporates connectivity and performance parameters is developed for computer networks as an alternative to conventional reliability or performance based measures. The proposed measure, called the average network throughput per unit installed capacity, measures the traffic that can be reliably transferred between any pair of nodes in a network with unreliable components per unit installed capacity. This measure is extended to distributed computing systems. The two new measures related to distributed systems are called average distributed program throughput per unit installed capacity and average distributed system throughput per unit installed capacity. An algorithm to evaluate these three measures has been developed. Examples are provided to illustrate the effectiveness of these measures over conventional reliability measures.<
>