S P H E R A Controlling networks while maintaining resilience Baruch Barzel 1
Challenges 5.5 ร 10 7 People affected 10 2 Fatalities 6 ร 10 9 USD in damages 2
Structure vs. dynamics Structural perturbation (component failure) Dynamic outcome Can we predict the (Resilience loss) point of Resilience loss?
Dynamic framework ๐ ๐๐ฆ ๐ ๐๐ข = ๐บ ๐ฆ ๐ ๐ข , ๐ ๐ + เท ๐ต ๐๐ ๐ ๐ฆ ๐ ๐ข , ๐ฆ ๐ ๐ข , ๐ ๐๐ ๐=1 ๐ ๐ ๐ State of a system component (node) โข Concentration of a protein/metabolite โข Probability of infection of an individual 4 โข Load on power/communications component
Dynamic framework ๐ ๐๐ฆ ๐ ๐๐ข = ๐บ ๐ ๐ฆ ๐ ๐ข , ๐ ๐ + เท ๐ต ๐๐ ๐ ๐๐ ๐ฆ ๐ ๐ข , ๐ฆ ๐ ๐ข , ๐ ๐๐ ๐=1 Interaction mechanisms Self dynamics ๐บ Interaction mechanisms ๐ Distributed parameters ๐ ๐ , ๐ ๐๐ ๐ฝ ๐๐ ๐ ๐ฆ ๐ ๐๐ฆ ๐ ๐พ ๐ + เท ๐๐ข = โ๐ท ๐ ๐ฆ ๐ ๐ต ๐๐ ๐ฝ ๐๐ ๐ ๐๐ + ๐ฆ ๐ ๐=1 5
Dynamic framework ๐ ๐๐ฆ ๐ ๐๐ข = ๐บ ๐ฆ ๐ ๐ข , ๐ ๐ + เท ๐ต ๐๐ ๐ ๐ฆ ๐ ๐ข , ๐ฆ ๐ ๐ข , ๐ ๐๐ ๐=1 Interaction mechanisms Network structure Example: population dynamics ๐ฉ ๐๐ ๐ฝ ๐๐ ๐ ๐ฆ ๐ ๐ฆ ๐ ๐๐ฆ ๐ 1 โ ๐ฆ ๐ ๐๐ข = ๐ฆ ๐ + เท ๐ต ๐๐ ๐ฝ ๐๐ ๐ท ๐ ๐ ๐๐ + ๐ฆ ๐ ๐ฆ ๐ ๐=1 6
Dynamic framework ๐ ๐๐ฆ ๐ ๐๐ข = ๐บ ๐ฆ ๐ ๐ข , ๐ ๐ + เท ๐ต ๐๐ ๐ ๐ฆ ๐ ๐ข , ๐ฆ ๐ ๐ข , ๐ ๐๐ ๐=1 Interaction mechanisms Network structure ๐ฉ ๐๐ โข Nonlinear Weighted โข Multi-parametric ( ๐ ๐ , ๐ ๐๐ ) Heterogeneous (Scale-free) โข Black-box: ๐บ ๐ , ๐ ๐๐ sometimes unknown 7
Diverse and unpredictable Diverse & unpredictable State Can we predict the point of Resilience State loss? State Universal resilience patterns in complex networks. Nature 530 , 307 (2016) 8
A physicists nightmare Current nonlinear dynamics theory: Where real networks are: ๐ spans Each node has ๐ = 6 nearest orders of magnitude neighbors โข Disordered and weighted โข Low dimensional โข Extremely heterogeneous โข Symmetric structures (lattice or lattice-like) โข Scale free: ๐ ๐ โผ ๐ โ๐ฟ 9
Symmetry Zero order symmetry ๐ -order symmetry Each node has ๐ = 6 nearest neighbors All nodes identical All environments identical 10
Global control parameter Diverse & unpredictable Universal Activity State State Universal parameter ๐พ eff universally predicts the critical transition points of ๐ธ ๐๐ ๐ Can we predict the resilience loss point of Resilience Activity State ๐ธ ๐๐ ๐ = ๐ โค ๐ฉ ๐ ๐ State loss? ๐ โค ๐ฉ๐ ๐ธ ๐๐ ๐ Activity State State ๐ธ ๐๐ ๐ Universal resilience patterns in complex networks. Nature 530 , 307 (2016) 11
Global control parameter Structure ๐ฉ ๐๐ Well mapped Control parameter Our prediction ๐ธ ๐๐ ๐ = ๐ โค ๐ฉ ๐ ๐ ๐ โค ๐ฉ๐ Translating Structure into Dynamic observables of interest Example : Using the Structure of the power network to determine its Dynamic resilience against Resilience local failures or load perturbations A Dynamic observable of the system that we seek to predict, understand and influence Universal resilience patterns in complex networks. Nature 530 , 307 (2016) 12
Top-down - Global control parameter ๐ ๐๐ฆ ๐ ๐๐ข = ๐บ ๐ฆ ๐ ๐ข , ๐ ๐ + เท ๐ต ๐๐ ๐ ๐ฆ ๐ ๐ข , ๐ฆ ๐ ๐ข , ๐ ๐๐ ๐=1 ๐ ๐ ๐ธ ๐ ๐ ๐ธ ๐ Bridges between Topology and Dynamics Sets guidelines for ๐ธ intervention Universal resilience patterns in complex networks. Nature 530 , 307 (2016) 13
Bottom-up - Intervention ๐ ๐๐ฆ ๐ ๐๐ข = ๐บ ๐ ๐ฆ ๐ ๐ข , ๐ ๐ + เท ๐ต ๐๐ ๐ ๐๐ ๐ฆ ๐ ๐ข , ๐ฆ ๐ ๐ข , ๐ ๐๐ + ๐ถ ๐๐ ๐ ๐ ๐ข ๐=1 Structural interventions Dynamic interventions ๐ป ๐ (๐) Removing External signals nodes, adding ๐ป ๐ (๐) to selected links, changing nodes weights Functional interventions Manipulating ๐ฎ ๐ and ๐น ๐๐ or their parameters 14
Spatio-temporal spreading patterns Diverse & unpredictable Universal Control parameter ๐พ ๐ฉ ๐๐ Individual node response time ๐พ ๐ ๐ โผ ๐ ๐ ๐บ ๐ ๐ฆ ๐ , ๐ ๐ ๐ ๐๐ (๐ฆ ๐ , ๐ฆ ๐ , ๐ ๐๐ ) Time Time 15 Predicting the spatio-temporal propagation of signals in complex networks. Nature Physics. Hopefully soon
Soft stability ๐พ < ๐ ๐พ = ๐ ๐พ > ๐ Beyond stability How much time do we have before an undesired transition occurs 16
Optimizing functionality vs. resilience Beyond stability How much time do we have before an undesired transition occurs 17
Open threads Top-down Identify macroscopic control parameters ( ๐พ, ๐ ) Bottom-up Selecting nodes for intervention (real-time mitigation) Stability vs. resilience Enriching the discussion on stability Functionality vs. resilience Can we introduce balanced incentives 18
In the top-down approach systems are influenced by means of global control parameters. Quite often these act as boundary conditions for the system dynamics. To identify such control parameters is a challenge on its own. Often they can be derived from the known macroscopic, or system dynamics. As a major conceptual drawback, control parameters usually reflect limitations of stability, rather than of resilience. โข In the bottom-up approach systems are influenced by specifically targeting some of the system elements, e.g. agents in an agent-based model or nodes in a network representation. Again, two different possibilities exist: (i) the agents can be controlled in their internal dynamics, or (ii) the agent interactions can be controlled. 19
Can we identify general principles for the bottom-up control of socio- economic or ecological systems? How can driver nodes be identified based on data-driven methods? Can we explain the breakdown of resilience in social organizations as a misallocation of resources? What is the relation between resilience and the natural tendency of systems to maximize their performance? 20
Recommend
More recommend