Architecture for Software-Defined Wireless Networks Julius - - PowerPoint PPT Presentation

architecture for software defined
SMART_READER_LITE
LIVE PREVIEW

Architecture for Software-Defined Wireless Networks Julius - - PowerPoint PPT Presentation

AeroFlux: A Near-Sighted Controller Architecture for Software-Defined Wireless Networks Julius Schulz-Zander, Nadi Sarrar, Stefan Schmid WiFi needs SDN! Network services as applications Channel Selection Coarse Grained Control


slide-1
SLIDE 1

AeroFlux: A Near-Sighted Controller Architecture for Software-Defined Wireless Networks

Julius Schulz-Zander, Nadi Sarrar, Stefan Schmid

slide-2
SLIDE 2
  • Network services as applications

– Channel Selection – Mobility Management – Client Load Balancing – WIDS/WIPS Systems – Traffic aware TX control

  • Per-flow TX settings
  • e.g., PHY rate, retry, power

WiFi needs SDN!

Local Visibility Frequent Events Fine Grained Control Coarse Grained Control Global visibility

slide-3
SLIDE 3

The Case for Near-Sighted Controller

  • Fine grained control

– Transmission settings on a per-slice, per-client, and per-flow level – Traffic and application aware transmission control – E.g., live video streaming optimizations through adaptive (Direct) multicast strategies

  • Approach: handle frequent, localized events close

to their origin

slide-4
SLIDE 4

SDN for WiFi

  • OpenFlow (OF) for the integration with wired part
  • OF enabled Access Points (APs)
  • Separate south-bound interface for WiFi
  • Global Controller based on Odin1
  • Near-Sighted Controller close to the APs
  • WiFi specific Agents on APs handle

– Light Virtual Access Points (per client state) – WiFi datapath transmission entries

1 L. Suresh, J. Schulz-Zander, R. Merz, A. Feldmann, and T. Vazao. Towards programmable enterprise WLANS with Odin. In HotSDN ’12.

slide-5
SLIDE 5

Architecture

  • Global Controller for coarse-grained control

– Services that require global visibility – Middlebox handling beneficial

  • Near-sighted controller for fine grained

control

– Latency-critical and load-intensive tasks – Per-Flow transmission settings – Transmission statistic collection

  • Radio Agent exposes all functionalities

– LVAP abstraction – Handles WiFi Datapath entries – Provide interface for statistic collection

slide-6
SLIDE 6

Use Case: Live Video Streaming

  • Live video multicast stream
  • Match through OF rule
  • Redirected to Middlebox
  • Differentiation

– Keyframes – Interframes

  • WiFi DP rule for each subflow
  • Transmit keyframes more robust
  • Challenge: Frame Aggregation
slide-7
SLIDE 7

Current State

  • Global Controller
  • Datapath Element tested
  • Proof of concept NSC
  • Near-Sighted Controller is work in progress
  • Deployment ongoing (Lab deployment)
  • Deployment in production network planned for

this Summer