A routing scheme for datagram and virtual circuit services in the MSN

Manhattan Street Network (MSN) is a directed mesh network designed for a metropolitan area. To satisfy multiservice application, a routing scheme is required to support both datagram and virtual circuit services in MSN. The authors present a simple routing scheme that fulfils the need. The proposed scheme is based on computing hop counters (HCs), and then routing message(s) using their values. The hop counter computation is performed whenever a message is originated at a node to obtain a preferred path. Forward and backward learning are useful concepts in HC computation. The proposed routing scheme then discerns a virtual circuit or datagram service for a message in the network. To study the routing efficiency and traffic adaptivity, the authors have simulated this scheme, and the results compare favorably with that of an existing technique.<>