An array based technique for routing messages in distributed double loop networks

In Chalamaiah and Ramamurthy (1998), the shortest path between two processors in a distributed double-loop network, G(n;+h,+1), was found by computing the shortest path in the 4 networks, G(n;+h,+1), G(n;-h,1), G(n;h,-1), and G(n;-h,-1), and then finding the best amongst these paths. In this paper we present an algorithm that is faster as it generates the path directly. Our algorithm is based on an array, and can be used for routing a message between processors in a distributed loop network via the shortest path. This algorithm fills in an array of size n (n is the number of nodes in the network) with an integer which is the shortest distance to node 0, in d time steps (where d is the diameter of the network). This is based on breadth-first search techniques (West 2001). Knowledge of the diameter is not required. The shortest path between two nodes can then be obtained using this array.