1 / 12

Breakout Session 3

This workshop explores new models and technologies in the physical layer of wireless networking, the tradeoff between realistic and tractable models, and the need for new techniques in algorithm theory and optimization. Topics include distributed optimization, fault-tolerant algorithms, and exploiting the structure of underlying problems. Relevant issues in networking and algorithm design are also discussed.

mcspadden
Télécharger la présentation

Breakout Session 3

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. NSF Workshop on Bridging Gap Between Wireless Networking Technologies and Advances in Physical Layer Breakout Session 3

  2. Participants • Mario Gerla, Michael Honig, Tom Hou, Vijay Kumar, Tom Luo, Narayan Mandayam, Madhav Marathe, Anna Scaglione, Kang Shin, R Srikant, Aravind Srinivasan, Tan Wong, Lizhong Zheng

  3. Summary • A number of advances are being made at creating new and more realistic models at PHY layer • E.g. DMT models that quantify reliability rate trade-off • New technologies in form of Cognitive Radios, MIMO radios and Cooperative communication will likely result in new PHY layer abstractions • Current algorithmic/optimization methods often do not use the new models • Tradeoff between realistic models and tractable models • New techniques in algorithm theory and combinatorial optimization need to be developed • Higher layers can benefit from an understanding of the PHY layer

  4. PHY layer models: 1 • Holistic view of control and data needs to be taken • New and more realistic models at PHY layer are now available • E.g. what is the networking viewpoint on the Diversity-Multiplexing Gain Tradeoff, for wireless multihop networks ? • Layered approach should be maintained to the extent possible; interfaces need to be created to expose information between layers for cross layer optimization • Distinguish between models versus metrics

  5. PHY Layer Models: 2 • New technologies such as cognitive radios are currently being developed. PHY level abstractions for these new technologies would provide the first step to more realistic protocol design and algorithmic analysis

  6. Issues in Optimization and Algorithms: 1 • Many problems are cast as mixed (continuous as well as discrete variables) non-convex programs • Very little work has been done in distributed optimization • Good working definitions of distributed algorithms • Robust Optimization: algorithms should work when instances vary slightly, proof techniques should work when models are varied slightly • Fault Tolerant algorithms

  7. Issues in Algorithms and Optimization: 2 • Algorithms and analysis based on simple interference models such as disk graphs should not be discarded off hand • Even though theorems might not extend, the proof techniques might still be extensible • Simple models are amenable to analysis and one needs to quantify the additional gains that can be made by using more complicated models • Generic algorithms: algorithms work for all successively complicated models, analysis depends on the model at hand

  8. Issues in Networking • Better systems and high level protocols can be built by understanding the PHY level abstractions • E.g. Cognitive radio helps reduce the interference and enrich topology options (such as TCP fairness) • New application areas such as Vehicular networks, Underwater networks, etc. require modifying PHY models • Spatial and temporal variation in usage of spectrum can be used for better spectrum utilization

  9. Summary • A number of advances are being made at creating new and more realistic models at PHY layer • E.g. DMT models that quantify reliability rate trade-off • New technologies in form of Cognitive Radios, MIMO radios and Cooperative communication will likely result in new PHY layer abstractions • Current algorithmic/optimization methods often do not use the new models • Tradeoff between realistic models and tractable models • New techniques in algorithm theory and combinatorial optimization need to be developed • Identifying classes of integer programs and non-convex programs that can be solved efficiently: exploiting the structure of the underlying problems

  10. What are we asked to Cover: 1 • Comments on talks so far • Specific Comments on Talks • Important issues that are overlooked • For physical layer folks: • what are other recent advances at physical layer, what are their impact on wireless networking, future • expected advances/breakthroughs at physical layer and how they will impact wireless networking research, where are the research gaps

  11. What are we asked to Cover: 2 • Wireless networking folks: • Current status (where we are now), future expectations on wireless networking and research • Challenges, desired technology advances/ breakthrough from the physical layer, new advances needed from theoretical perspective • Algorithm design and optimization: • Status and Open Problems • Advances/breakthrough at the physical layer, new challenging problems arising from future wireless networking, identify research gaps

  12. Models: • Realistic yet computationally tractable • Get provable bounds using realistic models • Recent techniques in optimization have not been exploited • Over head of managing additional parameters

More Related