1 / 18

Low-complexity Scheduling for Wireless Networks

Low-complexity Scheduling for Wireless Networks. Guanhong Pei ∗ , V. S. Anil Kumar †. ∗ Dept. of ECE and Virginia Bioinformatics Institute, Virginia Tech, Blacksburg, VA † Dept. of CS and Virginia Bioinformatics Institute, Virginia Tech, Blacksburg, VA. ACM MobiHoc 2012.

Download Presentation

Low-complexity Scheduling for Wireless Networks

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. Low-complexity Scheduling for Wireless Networks Guanhong Pei∗, V. S. Anil Kumar† ∗ Dept. of ECE and Virginia Bioinformatics Institute, Virginia Tech, Blacksburg, VA † Dept. of CS and Virginia Bioinformatics Institute, Virginia Tech, Blacksburg, VA • ACM MobiHoc 2012

  2. Wireless Network Model • The network is modeled as a graph • Set of nodes: • Set of transmission links: • Wireless Interference • Graph-based interference model • Interference set for each link • and : interfering ⇔no successful tx at the same time • Interference relationship • Binary & Symmetric Data → receiver link sender ← ACK

  3. Physical Interference Model • SINR • Signal to Interference & Noise Ratio • Condition for successful transmission on link : Interference from to : interference : Signal attenuation rate Signal SINR at receiver of Background Noise Interference from all other transmitting links Threshold

  4. Traffic & Queueing Models • Traffic: Single-hop • Exogenous arrival processes: general i.i.d. • : # arrival packets to link at time ; • First moment: • Queueing: Each link is associated with a queue • : # packets queued on link l at time t • : # arrival packets on link l at time t • : # departure packets on link l at time t • : service rate offered to link l at time t Queue Arrival Departure Service Queue Evolution:

  5. Queueing Stability • Long-term average backlog: • Queue-stability of a system iff • In a stable system • long-term average arrival rate equalslong-term average throughput rate

  6. Throughput Region Graph Interference model Scheduling & Power Assignment scheme SP Max-Weight Scheduling Throughput region: ΛSPthe set of all stable traffic vectors under SP Capacity region: ΛOPTthe set of all stable traffic vectors γ-scaled region γΛOPT, (0 ≤γ≤1) γ: efficiency ratio (0≤γ≤1)

  7. The Problem • Distributed • Low-complexity • Scheduling & Power control • SINR Interference • Performance • Important Issues & Techniques Maximize the efficiency ratio

  8. Related Issues • Overhead of Control • Ambiguity in the Distributed Model • Questions • Outdated& infrequently-updatedcontrol info? • Performance impact? Trade-off?

  9. Summary of Results Low-complexity Distributed Scheduling & Power Control:Adaptive Random-access Algorithm Based on Queueing Info RA-SCHED : Interference Degreemax # independent links in any interference set Graph-based Interference: • Throughput • Region: Physical Interference: Link Diversity# classes of link lengths RA-SCHED-SINR Efficiency Ratio Single-hop Setting: • General stochastic traffic Infrequent info-exchange Use stale queueing info updated infrequently Random-access based on queue size info Negligible overhead No additional overhead for exchanging control information Techniques: • Power control policy: simple and flexible

  10. Efficiency Ratio of RA-SCHED-SINR • Link Diversity • Worst-case Efficiency Ratio in Practice • Can be Ω(1) • Link diversity is small in practice: • In the most extreme cases: kilometerscentimeters max. link length min. link length

  11. RA-SCHED Frame i-1 Frame i • Frame i+1 Frame: Info-exchange sub-frame Scheduling-txsub-frame Backlog info exchange Each link transmits with Prob. at time Channel-access Probability

  12. RA-SCHED: Details Each link transmits with Prob. at time Backlog info exchange • Variables: • : Backlog of at the start of frame • : Sum of of links in • : • Backlog info exchange: 2 rounds • (1) distribute to all the links in • (2) distribute to all the links in • Channel Access Prob. • if • For every time slot in the scheduling-tx sub-frame of Interference set of Frame: Info-exchange sub-frame Scheduling-txsub-frame

  13. Analysis of RA-SCHED Length: H1 Length: H2 prob. of a successful tx in an interference set: close to 1/e at each time t ofscheduling-tx sub-frame Frame: Efficiency Ratio: Info-exchange sub-frame Scheduling-txsub-frame If total average arrival in each interference set: The total queue size in the network w.h.p.in some period of time Necessary Condition of Stability For any H1, H2, such that The network is queue-stable Sufficient Condition of Stability

  14. Schedule Augmentation Preparation for RA-SCHED-SINR • Partition link set L into g(L) length classes {Li} • T-augmented schedule Slot under Policy S Super-slot under Policy T-augmented S 1 2 3 T-1 T

  15. RA-SCHED-SINR • A g(L)-augmented version of RA-SCHED • Channel-access Probability • Redefine Interference Set • Power Assignment • Links in the same length class use “similar” power • Including uniform, square-root, linear assignments • Backlog Info-exchange : a constant to be determined during analysis : a constant

  16. Analysis of RA-SCHED-SINR • Same workflow as RA-SCHED • However, “interference sets” are conceptual • Extra work: links maketransmission attempts bad good non-localnon-linear causeinterference deliverpackets upper-bound on interference lower-bound onprob. of successful transmission : a constant to be determined during analysis connected & balanced by parameter:

  17. Analysis of RA-SCHED-SINR (cont.) • Incorporate g(L) into efficiency ratio • in our SINR Setting Efficiency Ratio: Efficiency Ratio:

  18. Conclusion • Low-complexity Distributed Algorithms • Scheduling & power control • SINR interference, K-hop interference • Efficiency ratios: and • Outdated & Infrequently-updated control info • Future Work • Multi-hop traffic • Heterogeneous periods of control info update • Quantify delay throughput trade-off • Adaptive CSMA in SINR setting

More Related