Supporting Demanding Wireless Applications with Frequency-agile - - PowerPoint PPT Presentation

Supporting Demanding Wireless Applications with Frequency-agile Radios

Lei Yang, Wei Hou*, Lili Cao, Ben Y. Zhao, Haitao Zheng

Department of Computer Science, University of California, Santa Barbara *Department of Electronic Engineering, Tsinghua University

1 NSDI 2010

Multimedia Streaming in Home/Office

2

• Real-time multimedia

flows in home/office networks

• High bandwidth
• QoS requirements
• No messy wires

NSDI 2010

Supporting Wireless Media Sessions

• Desired properties

– Continuous access to radio spectrum, high-bandwidth transmissions – Support multiple concurrent flows – Adapt to time-varying traffic demands

3 NSDI 2010

Can We Use WiFi?

• The 2.4G/5G ISM band is too crowded  no dedicated access
• Per-packet CSMA contention  frequent & unpredictable

disruptions

4

Time

NSDI 2010

Per-session Frequency Domain Sharing

• Simultaneous media sessions work in parallel on

isolated frequencies

5

Time Frequency

No interference Continuous spectrum access in time On-demand frequency usage

NSDI 2010

Is This Feasible?

• Opportunity for new dedicated

frequency band

– FCC has auctioned & released new spectrum – Start from a clean spectrum band

• Opportunity to deploy new access

mechanism

– The new National Broadband Plan encourages new dynamic spectrum access mechanisms

6 NSDI 2010

Our Design: Jello, per-session frequency domain sharing

Jello: Decentralized Home Media System

7 NSDI 2010

GNU Radio SORA AirBlue

Jello’s Key Components

8

Where is the usable spectrum?

Which frequency band should I use?

NSDI 2010

Spectrum Sensing

Identify available spectrum to avoid interfering with others

Spectrum Selection

Select spectrum to match traffic demands and utilize spectrum efficiently

Coordination

Sender receiver coordinate to calibrate sensing results and synchronize spectrum usage

How to Identify Free Spectrum?

• Conventional sensing: energy detection

– Simple, but unreliable

9

Threshold too high Threshold too low rising edge dropping edge Power Spectrum Density Frequency

Jello devices identify and use such edge patterns to get better sensing!

NSDI 2010

A unique feature: clear edge patterns on power spectrum density map

Sensing via Edge Detection

• Step 1: Preprocessing

– Smoothing by averaging over multiple observations

• Step 2: Detecting edges

– Calculate 1st order derivative of the power spectrum map – Identify rising/dropping edges

10

Power Spectrum Density Frequency 1st Order Derivative Frequency

Rising edges Dropping edges

Positive Negative

NSDI 2010

Much more robust than energy detection!

Choosing Frequency Blocks

11

Frequency

Usable frequency blocks New link

Our experiments reveal another fundamental challenge …

NSDI 2010

2.4G 2.5G

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.5 0.6 0.7 0.8 0.9

Disruption Rate Normalized Average Traffic Load

W/ Fragmentation W/o Fragmentation

Spectrum Fragmentation

• Like disks and memory, dynamic spectrum access

creates spectrum fragmentation

– Link comes and leaves – Link changes spectrum usage

12

Frequency Spectrum fragments

20% disruption

Link1 Link2 Link3

W/ Fragmentation W/o Fragmentation

NSDI 2010

4 video flows

Lead to significant media disruptions!

Link4

Solution 1: Defragmentation

• Rearrange global spectrum usage

– No, cannot stop all transmissions

• Our solution: individual online

defragmentation

– Voluntarily change spectrum usage to reduce fragmentation – Stays transparent to other links – Self-disruption

13

Frequency

Link1 Link2 Link3

Cannot eliminate fragmentation entirely, low levels of fragmentation may still exist

NSDI 2010

Defrag occurs infrequently

Solution 2: Non-Contiguous Spectrum Access

• Frequency-agile radios redesign

PHY to support non-contiguous spectrum access

– Combine multiple spectrum slices to form a single transmission – Decentralized OFDMA

• Fragmentation is no longer harmful
• Additional costs

– Increased frequency overhead

14

Link1 Link2 Link3

Non-contiguous frequency access reduces the impact of fragmentation, but at additional costs

Guard band

NSDI 2010

Frequency

A Unified Solution in Jello

Online defragmentation

• Remove most fragments
• Cannot completely remove

fragmentation

Non-contiguous access

• Effective for low fragmentation
• Increased frequency overhead

and hardware complexity

15

Frequency Link1 Link3 Frequency Link1 Link3

The two techniques are complementary to each other

NSDI 2010

Proof of Concept Implementation

• GNU Radio

– Software Define Radio – USRP frontend at 2.4G – Widely available, inexpensive, flexible

• Hardware limitations

– Limited bandwidth: 500kHz OFDMA – Large and unpredictable proc. delay

16

See the paper for detailed implementation

NSDI 2010

Spectrum Sensing Spectrum Selection Coordination

Testbed Evaluation

• 8-node GNU Radio testbed

– 4 concurrent flows – 12m x 7m room with various furniture and walls

• Traffic load

– Video and synthetic traces

17

• Evaluated 4 systems

– Static: Partition spectrum equally, WiFi-like – Jello-C: Jello with contiguous frequency access – Jello-Full: Full version of Jello – Optimal: Oracle solution w/o fragmentation and overhead

NSDI 2010

Overall Media Quality

Video Disruption Rate: percentage of time video is disrupted

18 NSDI 2010

0.1 0.2 0.3 0.4

0.5 0.6 0.7 0.8

Disruption Rate Normalized Average Traffic Load

Static Jello-C Jello-Full Optimal

0.1 0.2 0.3 0.4 0.5

0.5 0.6 0.7 0.8

Residual Usable Spectrum Normalized Average Traffic Load

Optimal Jello-Full Jello-C

Spectrum Usage Efficiency

Residual Usable Spectrum: the amount of spectrum a new link can use

19

45% more spectrum for a new session

NSDI 2010

Conclusions

• Jello: the 1st system supporting demanding wireless

media sessions

– Per-session frequency domain sharing – Detect available spectrum: Edge-detection spectrum sensing – Reduce spectrum fragmentation: Non-contiguous spectrum access + online defragmentation

• We deploy Jello on 8-node GNU Radio testbed

– Support 4 concurrent flows – Provide high utilization and adapt to dynamic demands

20 NSDI 2010

Questions?

Full Jello Implementation available @

• http://www.cs.ucsb.edu/~htzheng/papyrus/

Jello Demo available @

• http://www.cs.ucsb.edu/~htzheng/papyrus/detail/demo.html
• http://www.youtube.com/watch?v=-BcycTXh4uc

21 NSDI 2010