Social-Insect-Inspired Networking for Autonomous Load Optimisation

Abstract

The now established paradigm of System on Chip advances towards high-density many-core systems by employing Networks on Chip (NoCs) to connect a large number of processing elements. This brings many new challenges to embedded system design, but also opens up opportunities for new approaches that can leverage and exploit the many-core fabric in ways that traditional system architectures could not. This paper describes a new approach of adopting behavioural aspects of social insects as an inspiration towards autonomous, self-repairing systems. Each node in the NoC is considered as a member of a distributed colony and a simple adaptive controller is responsible for determining the behaviour of each node (member of the colony), relying only on a set of sensory inputs local to each node and small amounts of information shared between neighbours. Three adaptive routing schemes are presented that demonstrate how a sense-act-model applied at node level can lead to the emergence of novel routing behaviours that provide runtime self-optimisation of network load balancing without any prior analysis of the NoC topology. It is also discussed how these emergent behaviours of systems inspired by social insect colonies can be exploited to exhibit other novel autonomous and adaptive behaviours including fault tolerance and dynamic task allocation.