The Decentralised Coordination of Self-Adaptive Components for Autonomic Distributed Systems

摘要: ion Control signal system model Disturbance Inputs Outputs Services Information Infrastructure Adaptive Architecture Controller Figure 2.9: Hinnelund's of an autonomic computing as a control system. systems examples closed-loop (Taylor and Tofts, 2004). that cannot be controlled directly are called disturbance inputs. Relationships between inputs outputs can linear or non-linear, with being favoured over non-linear due to their predictability. State Monitoring A state-space used describe has more than one input and/or output. The size the should tractable. monitoring states is equivalent in terms soliciting feedback from operation order guide adaptive behaviour (Dutton et al., 1997). information usually satis es Markov property 1997), decision making components generally make decisions without recourse historical (Hinnelund, 2004; Dutton Techniques have been developed, such estimators observers handle situations where measurements state made must computed available measurement data. In (see gure 2.9) there no support for abstract state. He provides sample 2004) implements service work ow.