In Infrastructure for Enterprise Wir ireless Networks Vivek - - PowerPoint PPT Presentation

in infrastructure for enterprise
SMART_READER_LITE
LIVE PREVIEW

In Infrastructure for Enterprise Wir ireless Networks Vivek - - PowerPoint PPT Presentation

Vid idyut: Exploiting Power Lin ine In Infrastructure for Enterprise Wir ireless Networks Vivek Yenamandra and Kannan Srinivasan Motivation Increasing demand for wireless capacity Proliferation of BYOD in workplaces Data


slide-1
SLIDE 1

Vid idyut: Exploiting Power Lin ine In Infrastructure for Enterprise Wir ireless Networks

Vivek Yenamandra and Kannan Srinivasan

slide-2
SLIDE 2

Motivation

  • Increasing demand for wireless

capacity

  • Proliferation of BYOD in workplaces
  • Data Intensive applications: Video

Streaming, Teleconferencing, Surveillance etc.

  • Scare spectrum resources

Growing emphasis for spectrally efficient large capacity wireless networks

2

slide-3
SLIDE 3

Enterprise WLAN

Ethernet Backbone AP 1 2 3 4 Dense Client Distribution

3

slide-4
SLIDE 4

Enterprise WLAN

Ethernet Backbone AP 1 2 3 4 Dense Client Distribution

4

slide-5
SLIDE 5

Enterprise WLAN

Ethernet Backbone AP 1 2 3 4 Dense Client Distribution The APs share medium(time/frequency/code) to mitigate interference

5

slide-6
SLIDE 6

Alternative?

Ethernet Backbone 2 3 4 1 Multiple APs coordinate to emulate a single virtual AP with many antennas – Network MIMO

6

slide-7
SLIDE 7

Why Network MIM IMO?

Ethernet Backbone 2 3 4 1 All four APs can serve their clients simultaneously without needing to share the medium.

7

slide-8
SLIDE 8

Network MIM IMO Prerequisite

The coordinating APs need to be synchronized in frequency and time

8

slide-9
SLIDE 9

Cannot synchronize

Network MIM IMO Im Implementation1

Ethernet Backbone Lead AP Synchronization Header Synchronize 2 3 4 1

9

Cluster

[1]. Hariharan et.al, JMB: Scaling Wireless Capacity with User Demands, SIGCOMM, 2012

slide-10
SLIDE 10

Network MIM IMO Im Implementation

Ethernet Backbone 2 3 4 1 Frequency mismatch causes interference

10

slide-11
SLIDE 11

Network MIM IMO Im Implementation

Ethernet Backbone 2 3 4 1 The transmission range of the lead AP limits the number of APs that can coordinate to emulate a single large virtual AP

11

slide-12
SLIDE 12

How can we synchronize across clusters?

12

slide-13
SLIDE 13

Vidyut

Power lines Reference Clock transmitted on the Power Lines Each AP uses the reference clock on the power lines to synchronize their

  • wn carrier clocks using a PLL.

2 3 4 1

13

slide-14
SLIDE 14

Vidyut

Power lines Reference Clock transmitted on the Power Lines Each AP uses the reference clock on the power lines to synchronize their

  • wn carrier clocks using a PLL.

2 3 4 1

14

All APs are synchronized

slide-15
SLIDE 15

Vidyut

2 3 4 1 No Frequency mismatch = No interference

15

slide-16
SLIDE 16

Phase Locked Loop

Phase difference detector Phase Difference to Voltage Converter Low- pass filter VCO Fref Fo ÷ 𝑶 Feedback Path Distributed to the baseband clock, carrier clock Fo/N Reference clock from the power lines Power Line

16

slide-17
SLIDE 17

Phase Locked Loop

Phase difference detector Phase Difference to Voltage Converter Low- pass filter VCO Fref Fo ÷ 𝑶 Feedback Path Distributed to the baseband clock, carrier clock Fo/N Reference clock from the power lines Power Line

17

slide-18
SLIDE 18

Phase Locked Loop

Phase difference detector Phase Difference to Voltage Converter Low- pass filter VCO Fref Fo ÷ 𝑶 Feedback Path Distributed to the baseband clock, carrier clock Fo/N Reference clock from the power lines Power Line

18

slide-19
SLIDE 19

Phase Locked Loop

Phase difference detector Phase Difference to Voltage Converter Low- pass filter VCO Fref Fo ÷ 𝑶 Feedback Path Distributed to the baseband clock, carrier clock Fo/N Reference clock from the power lines Power Line

19

slide-20
SLIDE 20

Phase Locked Loop

Phase difference detector Phase Difference to Voltage Converter Low- pass filter VCO Fref Fo ÷ 𝑶 Feedback Path Distributed to the baseband clock, carrier clock Fo/N Reference clock from the power lines Power Line

20

slide-21
SLIDE 21

How to select the reference frequency?

21

slide-22
SLIDE 22

Selecting the Reference Frequency

  • Determined by the Power Distribution Network
  • Elements like transformers/distribution panels

22

slide-23
SLIDE 23

Measuring Characteristics

23

slide-24
SLIDE 24

24

Transformer Response

Primary Secondary

slide-25
SLIDE 25

Transformer Response

  • 20
  • 15
  • 10
  • 5

5 1 2 3 4 5 6 7 8 9 10 Gain(dB) Frequency(MHz) Same Phase

25

slide-26
SLIDE 26

Transformer Response

  • 20
  • 15
  • 10
  • 5

5 1 2 3 4 5 6 7 8 9 10 Gain(dB) Frequency(MHz) Same Phase

Filtering effect

26

slide-27
SLIDE 27

Three-Phase Power Supply

27

The three phases are physically isolated Do we need a separate reference clock for each phase?

slide-28
SLIDE 28

28

Transformer Response

slide-29
SLIDE 29
  • 20
  • 15
  • 10
  • 5

5 1 2 3 4 5 6 7 8 9 10 Gain(dB) Frequency(MHz) Same Phase Cross Phase

Transformer Response

29

slide-30
SLIDE 30
  • 20
  • 15
  • 10
  • 5

5 1 2 3 4 5 6 7 8 9 10 Gain(dB) Frequency(MHz) Same Phase Cross Phase

Transformer Response

Site of coupling across phases

30

slide-31
SLIDE 31
  • 20
  • 15
  • 10
  • 5

5 1 2 3 4 5 6 7 8 9 10 Gain(dB) Frequency(MHz) Same Phase Cross Phase

Transformer Response

Site of coupling across phases

31

We need just a single reference clock

slide-32
SLIDE 32

Evaluation: How effective is Vidyut’s phase synchronization?

32

slide-33
SLIDE 33

Evaluating Phase Mismatch

Both APs synchronized using Vidyut

Φmismatch = (Freceived – Fpilot ) x T + Φinitial

33

Freceived

slide-34
SLIDE 34

Evaluating Phase Mismatch

Both APs synchronized using Vidyut

Φmismatch = (Freceived – Fpilot ) x T + Φinitial

When both nodes are synchronized, Freceived = Fpilot making Φmismatch constant over time

34

Freceived

slide-35
SLIDE 35

Phase Synchronization Over Time

No deteriorating trend over time

35

slide-36
SLIDE 36

Phase Synchronization Over Time

The randomness is introduced by the phase noise in the PLL

36

slide-37
SLIDE 37

Phase Synchronization Over Time

We observe a phase mismatch under 0.05 radians over 90% runs.

37

slide-38
SLIDE 38

Power Distribution Network

  • Power lines are designed to carry power at 50/60 Hz
  • The higher frequency of the reference clock

attenuates over distance.

Each AP regenerates the reference clock back on to the power lines

38

slide-39
SLIDE 39

Clock Regeneration

2 1 Reference Clock Ref In Ref In Each AP feeds back a Reference clock phase matched to Ref In back on to the power lines.

39

slide-40
SLIDE 40

Clock Regeneration

2 1 Reference Clock Ref In Ref In Each AP feeds back a Reference clock phase matched to Ref In back on to the power lines. Enables synchronization of spatially distant APs

40

slide-41
SLIDE 41

Clock Regeneration

2 1 Reference Clock Ref In Ref In Each AP feeds back a Reference clock phase matched to Ref In back on to the power lines. Enables synchronization of spatially distant APs Makes Vidyut robust against single point of errors

41

slide-42
SLIDE 42

Regeneration Effect on Clock Synchronization

  • Each clock regeneration adds a distinctive phase noise characteristics
  • The phase mismatch between a pair of nodes does not correlate with

the number of clock regenerating sources between them.

  • Details in the paper.

42

slide-43
SLIDE 43

Achieving Distributed Time Synchronization

  • We adopt the principles proposed in [1].
  • Utilize the stable power frequency to achieve distributed time

synchronization

  • Details in the paper.

[1]. Rowe et.al, Low-power clock synchronization using electromagnetic energy radiating from ac power lines, SENSYS, 2009

43

slide-44
SLIDE 44

Im Implementation

  • Eight NI based SDR nodes
  • NI-5791 RF Front End
  • Accepts Reference Input/ Drives PLL output
  • 10 MHz OFDM in the 2.4 GHz ISM Band
  • PXIe-7965R FPGA ..
  • Agilent 8648C : 10 MHz Reference Clock

44

slide-45
SLIDE 45

Testbed

We interface the nodes to random power outlets across all three phases of power supply

45

32m 20m

slide-46
SLIDE 46

Evaluation: Performance gains of Vidyut-enabled Network MIMO.

46

slide-47
SLIDE 47

Setup

Divide the eight nodes into four APs and four clients. Place the nodes at random locations as before such that the APs are divided into two clusters Each cluster has clients to service Compared schemes: MegaMIMO, NEMOx1

47

[1]. Zhang et.al, Scalable Network MIMO for wireless networks, Mobicom, 2013

slide-48
SLIDE 48

48

NEMOx

Ethernet Backbone 2 3 4 1 Frequency mismatch causes interference

48

slide-49
SLIDE 49

NEMOx

Ethernet Backbone 2 3 4 1

49

slide-50
SLIDE 50

Throughput Gain

0.2 0.4 0.6 0.8 1 0.5 1 1.5 2 Fraction Throughput Gain MegaMIMO NEMOx

Absence of Cross Cluster Interference

50

slide-51
SLIDE 51

As the Number of f Clusters In Increases

MATLAB based simulation Account for increase in noise at each client due to phase mismatch between APs as their number increases. Provisions slackness for variance in time synchronization MegaMIMO and NEMOx are implemented using a TDMA over CSMA type MAC

51

slide-52
SLIDE 52

2 4 6 8 2 3 4 6 8 10 12 15 18 20 Throughput Gain Number of Clusters

MegaMIMO NEMOx

As the Number of f Clusters In Increase

Higher Density Low Density

52

slide-53
SLIDE 53

Future Work

  • Client selection in the clusters is an important design decision that has

been left for future work.

  • As the number of nodes participating in Network MIMO increases, the

challenge of processing the resulting large volumes of data needs to be addressed.

  • Distributed synchronization across multiple collision domains can enable

scalable implementation of exciting theoretical and systems work.

53

slide-54
SLIDE 54

Thanks!

54

Vidyut Language of Origin: Sanskrit Definition: Electricity Alternate Pronunciations: Probably will not help.