1 / 8

Decentralized Computing

Decentralized Computing. Group: Goce , Patrick, Claudia, Chris, Ouri, Robert, Markus, Moni. Goce ran off after review session …. And left me with these slides. General. “there is work already done in all relevant areas” (Ouri) – what is specific in context of CTS?

tammy
Download Presentation

Decentralized Computing

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Decentralized Computing Group: Goce, Patrick, Claudia, Chris, Ouri, Robert, Markus, Moni Goce ran off after review session …. And left me with these slides

  2. General • “there is work already done in all relevant areas” (Ouri) – what is specific in context of CTS? • Highly temporal data, rapidly changing information, heterogeneous data, data streams • Link between (distributed) analysis and control

  3. trucks cargo (ships) logistics car Aggregation Data; Information train Control bus Accuracy vs. Latency cycle walk Single location Central server Regional servers “D2D” (P2P; roadside) V2V Sequence Infrastructure No Infrastructure Full trajectory Uncertainty/Anonymization Global Overview of Issues in Transitions Among Contexts and Impact of Decentralization

  4. Possible Research Questions • Anonymization of Aggregates (full trajectories) • Background: Regulations, insert random objects, obfuscation/blurring • Methodologies/Tools to orchestrate and control the information flow (in/across infrastructure net) • Inventory of transportation tasks (w.r.t. decentralization, hierarchy, localization accuracy (precise position …. Daily plans) • Novel information-theoretic formalisms and measures for complexity/efficiency • Data freshness/staleness • Communication costs • Transition among the different levels of hierarchical decentralization • Current status / Background: • Kolmogorov Complexity in Wireless Sensor Networks; • Relational Complexity for Traditional Databases; • Triggers from Active Databases and more novel applications (TRG project)

  5. Possible Research Questions • Quality of Data guarantees (what are the trade-offs imposed by decentralization, transitioning across layers and the Quality of Service?); influence of local, i.e. usually only partial knowledge • Background: different error measures (probabilistic, fuzzy, …); • Quality measures for cross-layer execution of decentralized algorithms? • Background: Distributed Robots Coordination (Swarms); Data Integration/Quality • What constraints in CTS “force us” to use decentralized computing? Constraints Satisfaction in Decentralized Settings; • Categories of resource constraints (e.g. bandwidth, energy limits, computational capabilities in different contexts, …) • Dealing with Heterogeneous Contexts • Mobile networks • Background: Energy issues in Wireless Sensor Networks, Constraint Databases

  6. Possible Research Questions • Data Reduction + Aggregation in the context of streaming settings: • How to distribute the storage and management across hierarchical layers, where: • The structure of different layers evolves/changes over time (e.g., V2V ad-hoc networks); • Enrichment of granularities of information • The delegation of the computational tasks across the layers based on necessity vs. efficiency • Global: • Coupling of streaming data management and distributed control activities across the different layers in the hierarchy (Quality of Service vs. Quality of Data trade-off); • Traffic Flow Optimization in Decentralized Transportation Clusters (e.g., fleet mgmt “+” traffic controller”); • Current status/background: • Distributed Computing/Databases/ • Data Mining; • Swarm Intelligence; • Mobile/P2P Computing; • Data Integration/Quality; • Distributed Control; • Context-Awareness; • Views (Databases); • (Cartographic) generalization

  7. Possible Research Topics • Modeling and Representation of highly heterogeneous data • reconciliation in different contexts • Interpretation* • Ontologies, Semantic Interoperability (dynamic learning) • New categories (e.g., large-V2V-networks) emerge and vanish + interoperability managed dynamically. • Data Warehousing* • Rollup/Drill_Down in hierarchical decentralized settings Current Status/Background: - Ontologies/S.O./SOA - Business and Spatial Data Warehousing *NOTE: subject to discussion with the KD-group

  8. Comments from MixAndMatch-group • Different epistemic levels of P2P to substitute the current “regional servers” concept; • Delegating/Hand-off-ing the responsibility of “boss”: • Information transfer among nodes • In case of loss of “boss” – what is effect? • Basic assumptions about computational capabilities in different contexts (vehicles, pedestrians, road-side sensors); • Is there some “global” virtual machine • Restrictions on (de)centralizations/distribution(s) of non-technical/political/institutional/sponsorship nature. • Example(s) of entities involved in last bullet of Slide #4 • Generalizing the issues related to DW/DM outside the transportation world • Decision making, predictions

More Related