Locality and Efficiency in Distributed Computing

A prominent challenge is improving the efficiency of distributed computation: reducing the cost, load, and resource requirements while still maintaining the same quality of service. Due to the shear amount of information and the size of the network, executing global system-wide operations becomes very expensive. When possible, we wish to minimize the work done by the system and avoid the heavy cost incurred by global operations. One of the main tools to achieve efficient distributed computation is to exploit the issue of locality. Specifically, some tasks require the participation of a small region of the network. In such a case one would want the computation to span only the minimal required region. Routing is a prominent example of a task that can dramatically benefit from exploiting locality. If the source and destination nodes are close by, routing should involve only the small region that contains the source and destination. One of the fundamental trade-offs in compact routing schemes is between the space used to store the routing table on each node and the stretch factor of the routing scheme – the maximum ratio over all pairs between the cost of the route induced by the scheme and the cost of a minimum cost path between the same pair. A name-independent routing scheme is a distributed algorithm that allows any source node to route messages to any destination node, given the destination’s network identifier. This thesis suggests new ways in which distributed routing schemes can exploit locality in an efficient manner by proposing new and improved space-stretch tradeoffs for compact name-independent routing schemes.