Bartendr: A Practical Approach to Energy-aware Cellular Data - - PowerPoint PPT Presentation

▶
bartendr a practical approach to energy aware cellular
SMART_READER_LITE
LIVE PREVIEW

Bartendr: A Practical Approach to Energy-aware Cellular Data - - PowerPoint PPT Presentation

Jan 07, 2023 277 likes •612 views

Bartendr: A Practical Approach to Energy-aware Cellular Data Scheduling Aaron Schulman Vishnu Navda Neil Spring Ramachandran Ramjee Calvin Grunewald Venkata N. Padmanabhan University of Maryland Microsoft Research India Kamal Jain

slide-1
SLIDE 1

Bartendr: A Practical Approach to Energy-aware Cellular Data Scheduling

Aaron Schulman Neil Spring Calvin Grunewald University of Maryland Vishnu Navda Ramachandran Ramjee Venkata N. Padmanabhan Microsoft Research India Pralhad Deshpande Stony Brook University Kamal Jain Microsoft Research Redmond

slide-2
SLIDE 2
  • 130
  • 120
  • 110
  • 100
  • 90
  • 80
  • 70
  • 60
  • 50

20 40 60 80 100 120 140 160 signal strength (RSSI) 100 meter steps

20 40 60 80 100 120 140 160

2

A moving phone experiences signal strength variations

slide-3
SLIDE 3

Signal strength affects radio power and throughput

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.5 1 1.5 2 CDF throughput (Mbit/s) signal

  • 50
  • 60
  • 70
  • 80
  • 90
  • 100
  • 110
  • 120

500 1000 1500 2000 2500

  • 100 -90 -80 -70 -60 -50 -40 -30

power (mW) signal strength (RSSI)

3

slide-4
SLIDE 4

Signal strength affects radio power and throughput

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.5 1 1.5 2 CDF throughput (Mbit/s) signal

  • 50
  • 60
  • 70
  • 80
  • 90
  • 100
  • 110
  • 120

500 1000 1500 2000 2500

  • 100 -90 -80 -70 -60 -50 -40 -30

power (mW) signal strength (RSSI)

3

slide-5
SLIDE 5

Signal strength affects radio power and throughput

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.5 1 1.5 2 CDF throughput (Mbit/s) signal

  • 50
  • 60
  • 70
  • 80
  • 90
  • 100
  • 110
  • 120

500 1000 1500 2000 2500

  • 100 -90 -80 -70 -60 -50 -40 -30

power (mW) signal strength (RSSI)

3

slide-6
SLIDE 6

Signal strength affects radio power and throughput

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.5 1 1.5 2 CDF throughput (Mbit/s) signal

  • 50
  • 60
  • 70
  • 80
  • 90
  • 100
  • 110
  • 120

500 1000 1500 2000 2500

  • 100 -90 -80 -70 -60 -50 -40 -30

power (mW) signal strength (RSSI)

3

slide-7
SLIDE 7

Energy efficiency can be improved

A moving phone experiences signal strength variations. Signal strength affects communication energy. Applications can hold off until signal increases and prefetch while signal is strong.

4

slide-8
SLIDE 8

Energy efficiency can be improved

A moving phone experiences signal strength variations. Signal strength affects communication energy. Applications can hold off until signal increases and prefetch while signal is strong.

4

Bartendr

slide-9
SLIDE 9

Applications can receive when signal is strong

Background sync - 5 min interval sync could be more efficient if done sometime between 4 to 6 min Streaming media - Consume buffer when the signal is weak, prefetch when the signal is strong

5

slide-10
SLIDE 10

Application energy measurements

Drove with a mobile power monitor connected to a Palm Pre

6

slide-11
SLIDE 11

Email sync energy consumption

5 10 15 20

  • 105 -100 -95
  • 90
  • 85
  • 80 -75
  • 70
  • 65
  • 60 -55
  • 50

energy (J) signal strength (RSSI)

7

slide-12
SLIDE 12

Email sync energy consumption

5 10 15 20

  • 105 -100 -95
  • 90
  • 85
  • 80 -75
  • 70
  • 65
  • 60 -55
  • 50

energy (J) signal strength (RSSI)

✓ ✖

7

slide-13
SLIDE 13

YouTube energy consumption

~ ~

286 20 40 60 80 100 120 140 160 180

  • 93
  • 73

energy (J) signal strength (RSSI)

8

slide-14
SLIDE 14

Applications must schedule communication

Problem

When to schedule communication to save energy?

9

Predict signal strength Schedule syncs Schedule streaming Sync Streaming Schedule wakeup Fill the buffer efficiently

slide-15
SLIDE 15

Applications must schedule communication

Problem Challenge

Scheduling must save more energy than it consumes. When to schedule communication to save energy?

9

Predict signal strength Schedule syncs Schedule streaming Sync Streaming Schedule wakeup Fill the buffer efficiently

slide-16
SLIDE 16

Obstacles to energy efficient scheduling

energy consumer consumption Bartendr

Signal prediction locating the phone

  • n a path

(1D not 2 or 3D)

GPS is 400 mW and slow to fix, WiFi must be in receive mode

phone already maintains signal strength, cell id, and neighbor cells Sync scheduler wakeup and sleep

1 J to wake up 0.5 J to sleep schedule syncs minutes into the future

Streaming scheduler radio energy tail

3 - 10 s of radio power after communication (at least 400 mW) consider the radio’s power state when scheduling a stream

10

slide-17
SLIDE 17
  • 130
  • 120
  • 110
  • 100
  • 90
  • 80
  • 70
  • 60
  • 50

20 40 60 80 100 120 140 160 signal strength (RSSI) 100 meter steps 1 2 3 4 5 6 6 ... 1

20 40 60 80 100 120 140 160

Signal strength variation on a path

11

slide-18
SLIDE 18
  • 130
  • 120
  • 110
  • 100
  • 90
  • 80
  • 70
  • 60
  • 50

20 40 60 80 100 120 140 160 signal strength (RSSI) 100 meter steps 1 2 3 4 5 6 6 ... 1

20 40 60 80 100 120 140 160

Signal strength variation on a path

11

5

slide-19
SLIDE 19

Predicting signal strength with previous drives

  • 1. Find location in a previous drive

Signal strength, cell id, neighbor list

  • 2. Look ahead for future signal strength

12

seconds in the future

slide-20
SLIDE 20

Predicting signal strength with previous drives

  • 1. Find location in a previous drive

Signal strength, cell id, neighbor list

  • 2. Look ahead for future signal strength

12

seconds in the future

slide-21
SLIDE 21

Scheduling when to sync

Wake-up, sync, schedule, sleep Uses threshold for efficient sync Schedules for either first or widest signal

5 10 15 20

  • 105 -100 -95
  • 90
  • 85
  • 80 -75
  • 70
  • 65
  • 60 -55
  • 50

energy (J) signal strength (RSSI)

✓ ✖

13

slide-22
SLIDE 22

Scheduling when to sync

Wake-up, sync, schedule, sleep Uses threshold for efficient sync Schedules for either first or widest signal

13

first widest

  • 100
  • 95
  • 90
  • 85
  • 80
  • 75
  • 70
  • 65
  • 60

50 100 150 200 250 signal strength (RSSI) time (s)

slide-23
SLIDE 23

Scheduling when to receive a stream

Challenge

  • 1. Tradeoff between strong signal and radio tail energy
  • 2. Signal prediction error due to speed variations
  • 3. Throughput prediction error due to congestion

Approach

  • 1. Minimize predicted energy - dynamic programming algorithm
  • 2. Update schedule with latest signal prediction
  • 3. Schedule based on remaining buffer

14

slide-24
SLIDE 24

15

future past now

slide-25
SLIDE 25

Evaluation methodology

Simulated energy consumption of naive and scheduled syncs and streaming Several 17 km drives of throughput and signal for prediction and simulation of energy consumption Started at many points in the drive

16

slide-26
SLIDE 26

Syncing simulation

0.7 0.8 0.9 1 120 240 fraction of naive energy forced delay (s) 120 240 360 480 600 prediction window (s)

  • ptimal

first widest

17

ideal

slide-27
SLIDE 27

Syncing simulation

0.7 0.8 0.9 1 120 240 fraction of naive energy forced delay (s) 120 240 360 480 600 prediction window (s)

  • ptimal

first widest

17

ideal

slide-28
SLIDE 28

Syncing simulation

0.7 0.8 0.9 1 120 240 fraction of naive energy forced delay (s) 120 240 360 480 600 prediction window (s)

  • ptimal

first widest

17

ideal

slide-29
SLIDE 29

Syncing simulation

0.7 0.8 0.9 1 120 240 fraction of naive energy forced delay (s) 120 240 360 480 600 prediction window (s)

  • ptimal

first widest

17

ideal

slide-30
SLIDE 30

Streaming simulation

0.2 0.4 0.6 0.8 1 120 240 360 480 600 fraction of naive energy stream length (s) 64 kbit/s 128 kbit/s

18

slide-31
SLIDE 31

Related work

Breadcrumbs (A. J. Nicholson et al.) Predicts WiFi network quality for a mobile device Experiences in a 3G Network (Liu et al.) and An empirical study on 3G network capacity and performance (Tan et al.) Long term throughput at a location varies TailEnder (N. Balasubramanian et al.) and Cool-Tether (A. Sharma et al.) Batching and prefetching reduce radio energy tail

slide-32
SLIDE 32

Last call

Signal strength affects energy consumption Applications like sync and streaming can improve energy efficiency by deferring and prefetching Previous drives can predict signal strength without breaking the energy bank Scheduling can reduce energy consumption by up to 50% for large workloads and 10% for small

20