Wireless LAN Radio Spectrum Management Best Practices

Wireless LAN Radio: Spectrum Management Best Practices
BRKEWN-3013

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

1

What We’re Going to Cover –
  The Challenge   Wireless Trends
Evolution of the WLAN

  Deploying with Spectrum in Mind
Site Survey – A Word About Tools Cisco Radio Resource Management—RRM ClientLink, BandSelect, CleanAir

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

2

The Challenge

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

3

The Dynamic Nature of Spectrum
  You are breathing the physical layer   RF reflects off things   RF is absorbed by things   It’s a shared medium (as such, not all RF is always yours)   Requirements change in response to changes in the environment— not always helpful   Yet, if implemented and maintained properly, it’s a technology enabler providing
Increased productivity Creative freedom Enhanced user experience—by putting the power of the network where the user lives and works

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

4

Mobility Refers to the Client— Not the Infrastructure
  Radio assets are fixed devices   Autonomous AP channel and power must be set in advance   Clients move about   Resource demands shift with client location, and density
Clients Associate to AP with Strongest Signal

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

5

Even When Well Planned, Things Change
  Mission critical requires HA   Client technology refresh—additional device types   PDA’s, Tablets   New neighbors?

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

6

Does Non–Wi-Fi Interference Matter?
A Series of Papers on Wi-Fi Interference Concluded…
http://www.cisco.com/en/US/products/ps9393/prod_white_papers_list.html

Data
Normal Range Reduced Range

Voice

Video

Degraded Range

Reduced Coverage from 20% to 80%

Reduced Call Quality

Most Video Rated “Unwatchable”

..That Dramatic Loss in Quality of Mobility Services Will Result When Wi-Fi Encounters Interference
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

7

Wireless Trends

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

8

Enterprise Wireless Evolution
From Best Effort to Mission Critical

Hotspot

System Management

Scalable Performance

Self Healing & Optimizing

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

9

“When the students returned this year, if you asked me what percentage of students are using the Wi-Fi network – I would have told you 40%. I was shocked to see 85% of them using the Wi-Fi network.”
Scott Ksander – September 2009 – Cisco Education TAB Purdue University

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

10

Expectation for Mission Critical Wireless
Can’t “ IWithoutDo My Job Wireless. It Has to Work. ” Is Best“ WirelessCan’t Effort. I

Support a Level 1 SLA.

”

Continued Growth and Reliance on Wi-Fi Devices
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved.

vs.
Cisco Public

IT Lacks RF Resources and Expertise
11

Deploying with Spectrum in Mind

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

12

Deploying with Spectrum in Mind
  Role of site survey is as important as ever—but has evolved   Evaluate the existing application requirements, available spectrum and Clients   Focus should be on fixed infrastructure
AP placement Density is important Protocols supported Rates supported Interference sources

  Mitigating issues   Planning tools Designing for Sustainable Spectrum Management
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

13

A Word About Tools
How to Compare Apples to Apples   What you use is less important than how you use it   Internal vs. external adapters
Internal adapters – even the same model will have different antenna arays and placement for different model laptops External adapters – can be moved with the application – and provide consistent results – regardless of the platform used

  Use the same Tool to compare results!   Recheck results from a known environment with version updates   Free Tools – Nothing is Free
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

14

The Impact of a Crowded Spectrum

Performance at Risk in Unprotected Networks
Throughput Reduction Near
(25 Feet)

Interference Type

Far
(75 Feet)

End User Impact   Reduced network capacity and coverage   Poor quality voice and video   Potential complete link failure IT Manager Impact   Potential security breaches   Support calls   Increased cost of operation
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved.

2.4 or 5 GHz Cordless Phones Video Camera Wi-Fi
(Busy Neighbor)

100% 100% 90% 63% 20% 18%

100% 57% 75% 53% 17% 10%
15

Microwave Oven Bluetooth Headset DECT Phone Source: FarPoint Group
Cisco Public

What Is CCA and SOP?
  802.11 is CSMA/CA – collision avoidance   CCA is Clear Channel Assessment – and is the listen before talk component of Collision Avoidance   With 802.11n radios CCA is typically linked to Preamble/Start of packet   Radios are better these days (mostly)   CCA - is -65 and SOP is -85 dBm for 802.11b/g/a   If you can hear it above these levels – you are sharing the spectrum

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

16

CCA Blocked or High

802.11n Traffic Video Camera Duty Cycle 90-100% Video Signal

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

17

How Does Interference Impact Wi-Fi?
Separating the FUD from the Facts
  Collisions - Non Wi-Fi devices do not participate in our CA mechanism – they have their own rules   No respect for Wi-Fi – results in:
Corrupted packets Increased retransmissions Increased Duty Cycle Less available bandwidth

  SNR – Signal to Noise ratio

High SNR

Low SNR

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

18

802.11 and Duty Cycle – Channel Utilization
  Retransmit a packet   Duty Cycle of interference is logarithmically proportionate to channel time available   Busy network – less interference tolerance   Less busy – might not even notice low levels of interference   Bandwidth is like Money – the more you get the more you spend

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

19

Deploying with RRM in Mind

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

20

RRM—Radio Resource Management
  What are RRM’s objectives?
To dynamically balance the RF infrastructure and mitigate changes Monitor and maintain coverage for all clients Manage Spectrum Efficiency so as to maintain the optimal throughput under changing conditions

  What RRM does not do
Substitute for a site survey Correct an incorrectly architected network Manufacture spectrum

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

21

RRM Monitors the RF Group
  Continuously monitors dynamic changes in environment
Collection of statistics and metrics used by DCA, TPC, and CHDM Provides assessment of the overall “RF health” of the network

  Stats/metrics include:
Noise (e.g., radar, Bluetooth devices, microwave ovens) Interference (802.11—rogue APs) Signal – (our AP’s) Load

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

22

How Does RRM Do This?
  DCA—Dynamic Channel Assignment
Each AP radio gets a transmit channel assigned to it Changes in “air quality” are monitored, AP channel assignment changed when deemed appropriate (based on DCA cost function)

  TPC—Transmit Power Control
Tx Power assignment based on radio to radio pathloss TPC is in charge of reducing Tx on some APs— but may also increase Tx by defaulting back to power level higher than the current Tx level

  CHDM—Coverage Hole Detection and Mitigation

Detecting clients in coverage holes Deciding on Tx adjustment (typically Tx increase) on certain APs based on (in) adequacy of estimated downlink client coverage

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

23

Case Study 1 – College High Rise Dorm Channel Utilization
• 26 story dormitory • Low user count – but high channel utilization • Did an Active site survey • Customer complaint – disconnects and low throughput

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

24

WCS – Channel Utilization Report

0 TX and 0 RX Utilization – Channel 40-70%?

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

25

Duty Cycle – and Spectrum Capacity
  Duty Cycle is the on time of a given transmitter   It is measured as percentage of total time available, this relates directly to channel utilization, but is only part of the story – protocol overhead is the full story   802.11 can only do essentially two things to recover in a challenging RF environment
Retransmit a Frame – Turn the radio on again to send information that has already been sent once = Increased Duty Cycle Rate shift to a slower speed that can be supported – If retries are excessive, then the link will be rate shifted to a slower speed in an attempt to gain reliability

  Both of these will increase Duty Cycle and make the problem worse if it is a dense network

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

26

Understand Protocol Selection 802.11 b/g/a/n and Duty Cycle—Important? Why?
20000 18000 16000 14000 12000

CCK

DSSS

OFDM 64 Byte 128 Byte 256 Byte 512 Byte 1024 Byte 2048 Bytes

Time/µS

10000 8000 6000 4000 2000

Frame Size/Bytes
0 Mbps 1 2 5.5 11 6 12 24 36 48 54 130 300

Spectrum Is a Shared Finite Resource
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

27

Duty Cycle and Spectrum 802.11 b/g
Healthy Network
20-30% Duty Cycle Channel Separation

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

28

Duty Cycle and Spectrum 802.11 b/g
Unhealthy Network
No Channel Separation

100% Duty Cycle

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

29

Channel Utilization— What Made the Difference?

What Made This Dramatic Change?

Before 5% After

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

30

Every SSID Counts!

  Each SSID requires a separate Beacon   Each SSID will advertise at the minimum mandatory data rate   Disabled – not available to a client   Supported – available to an associated client   Mandatory – Client must support in order to associate

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

31

Cell Size – By Protocol/Speed
Assuming 10% PER
Speed 1 2 5.5 6 11 12 24 36 48 54 Required SNR 0 3 6 2 9 6 11 13 17 19 AP Sensitivity -91 -91 -91 -87 -88 -86 -85 -85 -78 -77

Channel Utilization – Is the Aggregate of Every Radio on the Channel That Can Be Heard Above -85 dBm – This Means Clients Too
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

32

RF Grouping Neighbor Messages = OTA – Over The Air - RF Analysis
  Neighbor AP messages are sent every 60 seconds at highest power and lowest supported data rate   Neighbor Messages are used by receiving APs and their WLCs to determine how to create inter and Intra-WLC RF Groups and Physical RF Neighborhoods   Each AP listens for other AP’s neighbor messages – and if it’s RF Group name matches – the message is forwarded to it’s controller and ultimately to the RF Group leader   A list is maintained for each AP in the RF Group of who heard his neighbor messages and how loud
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

Neighbor Messages Are Sent from Each AP to Multicast Address 01:0B: 85:00:00:00

33

Noise, Interference, and Utilization via WLC

RX Utilization 36 TX Utilization 7 Channel Utilization 96

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

34

WLC Config Analyzer View

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

35

RF Summary – Imbalance Between Neighbors

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

36

WCS – Map View – Show Neighbors
  Select any AP on the map and right click   Select View RF Neighbors   Table displays (1) Neighbors on the current map   Table displays (2) neighbors not on the current map

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

37

Three APs Deployed in Each Foyer

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

38

Initial Measures – Before and After
  Eliminated center 2.4 GHz radios – on each floor   Eliminated all but 11 Mbps   Enabled Client Link   Enabled Band Steering

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

39

End Result – APs Moved

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

40

Case 2 – RF Groups
  After conducting a multi floor active site survey using a 4400 and 10 x 1140 AP’s, coverage looked good at power levels 2-3.   The customer then deployed 3500 series AP’s according to the plan, and RRM set the power levels to 6!   What’s different about the 3500?

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

41

RF Grouping

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

42

RF Grouping

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

43

RF Grouping

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

44

The Tell… Survey Was Conducted in Separate RF Group

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

45

About RF Groups
RF Groups Are Clusters of Controllers that Share the Same RF Group Name. RF Neighborhoods Are Groups of APs that “Hear” Each Other
Wireless Controller A RF Group = <asciii string>

RF Group Controllers Elect an RF Group Leader That Analyses RF Data and Neighbor Relationships to Make More Intelligent Decisions About Optimizing the RF Environment for the System
Wireless Controller B RF Group = <ascii string>

> - 80dBm

Neighbor Messages Are Sent At Full Power, Containing Information About the APs Seen, and Authenticated via a MIC Based on the RF Group Name
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved.

IF APs on Different Controllers Hear Neighbor Messages from APs in the Same RF Group at –80 dBm or Greater They Will Group in an RF Neighborhood, Channel, and Power Then Compute as a Group
Cisco Public

46

RF Grouping and RF Neighborhoods
RF Neighborhood (a)

RFGroup - Bob

RF Neighborhood (b)

  Multiple “RF Neighborhoods” can exist within a single RF Group   RRM is calculated on a per RF neighborhood basis   RF Neighborhoods can be inter-controller or intra-controller   Multiple RF Neighborhoods may be formed even when controllers share same RF Group name
RF Groups/Neighborhoods Apply per PHY Type
RFGroup 1 RFGroup 1

RF Neighborhood (A)

RF Neighborhood (B)

RFGroup - Bob RF Neighborhood (E) RF Neighborhood (C) RF Neighborhood (A) RF Neighborhood (D) RF Neighborhood (B) RF Logical Neighborhood (C) RF sub-group (c)

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

47

Configuring RF Grouping
RF Group Name Is Configured From:   Controller > General on the WLC GUI:   Configure > Controllers > controller > System > General in WCS:

By Default the RF Network Name and Mobility Domain Name Are the Same, but This Is Default Behavior

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

48

RF Grouping
  By looking at the RF neighborhoods from the network perspective, you can determine which APs are literally within the same RF domain or neighborhood.   Placing like groups of APs into a separate RF group is perfectly ok, and in fact can provide much better design options

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

49

Case -3 DCA
  New Building installation   CU has a very high density of I-phone’s   Main Architect – good RF knowledge   Without RRM – channel distribution matched plan   With RRM – AP’s on same channel adjacent to one another   Did not trust RRM
Channel Utilization – vs. Spectrum Expert – did not match Neighbor Lists and spot check with Client card – vastly different

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

50

Neighbor Message and AP Neighbors
Did Not Trust RRM – Compared the Neighbor Lists for WLC to Beacon Observations at the AP Produced Less Trust -

Neighbor Messages Are Sent Every 60 Seconds to the Multicast Address of 01:0B:85:00:00:00
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

51

RRM Put Adjacent APs on the Same Channel!
  Looking at the 1st floor we see two APs on the same channel   At the 2nd floor, we see 3 APs   The 3rd, we see 3 APs   And the 4th we see 2 APs   But look at the APs channels as they stack! 11 1 6 1

1st

2nd

3rd

4th

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

52

WLCCA View

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

53

AP Placement
  Omni Antenna’s have an Elevation pattern of a donut   12 dB attenuation between floors   Customer intentionally stacked the AP’s to protect against direct exposure   Had these been 1130’s – possibly a valid argument

BUT - These Are Cisco AP 1140s

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

54

Antenna Pattern Elevation Plane
2.4 GHz

Access Point Has 3 Integrated 4 dBi (2.4GHz) Antennas

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

55

DCA 6.0 and Beyond
  CM= RSSI, Noise, Interference, signal, and a constant (threshold)   An AP list ordered by CM’s in the RF Neighborhood is created worst to best   Prior to release 6.0 – we solved for the worst AP CM in the RF Neighborhood   6.0 and after - DCA now operates on multiple local searches – and randomly selects CPCI (channel Plan Change Initiators) from the CPCI list and calculates optimal solutions for the CPCI and it’s first and second hop neighbors   The calculation completes with the NCCF function – a goodness value for the group – indicating positive change for the CPCI and it’s immediate neighborhood

Version 4.1.185.1 6.0

Band 2.4 GHz 5 GHz 2.4 GHz 5 GHz

High 5 dB 5 dB 5 dB 5 dB

Medium 15 dB 20 dB 10 dB 15 dB

Low 30 dB 30 dB 20 dB 20 dB

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

56

DCA Solution Flow
AP
AP-5 AP-4 AP-6 AP-1 AP-7 AP-8 AP-23 AP-14 AP-13 AP-19 AP-24 AP-25 AP-16 AP-10 AP-15 AP-17 AP-2 AP-11 AP-20 AP-21 AP-22 AP-9 AP-18 AP-3 AP-12

AP-16

CM
25 34 55 60 63 67 68 71 73 75 76 77 78 79 79 81 82 82 83 83 84 85 87 90 91

Worst

A CPCI List Is Created of All APs AP-11 in the Local RF Neighborhood AP-15
AP-19

AP-22

AP-6

AP-5

AP-5

AP-9

AP-18

AP-10
AP-21

Best
CPCI and First Hop Neighbor, Channel Change Is Allowed The Impact on the Second Hop Neighbor Is Considered in the Calculation, but No Channel Change Is Permitted
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

CPCI First Hop Neighbor Second Hop Neighbor
BRKEWN-3013

57

DCA Solution Flow
AP-16
AP
AP-5 AP-4 AP-6 AP-1 AP-7 AP-8 AP-23 AP-14 AP-13 AP-19 AP-24 AP-25 AP-16 AP-10 AP-15 AP-17 AP-2 AP-11 AP-20 AP-21 AP-22 AP-9 AP-18 AP-3 AP-12

CM
25 34 55 60 63 67 68 71 73 75 76 77 78 79 79 81 82 82 83 83 84 85 87 90 91

Worst

AP-11
AP-19

AP-15 AP-5
AP-5

AP-18

AP-6

AP-9

AP-21

AP-10
AP-22

Best
NCCF Is Calculated on the Entire Group for Each Channel Plan Calculated – A Plan Is Selected
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

CPCI First Hop Neighbor Second Hop Neighbor
BRKEWN-3013

58

DCA Solution Flow
AP
AP-5 AP-4 AP-6 AP-1 AP-7 AP-8 AP-23 AP-14 AP-13 AP-19 AP-24 AP-25 AP-16 AP-10 AP-15 AP-17 AP-2 AP-11 AP-20 AP-21 AP-22 AP-9 AP-18 AP-3 AP-12

CM
25 34 55 60 63 67 68 71 73 75 76 77 78 79 79 81 82 82 83 83 84 85 87 90 91

The CPCI – and Its First Hop Neighbors Are Removed from the CPCI List

Worst

AP-11

X

AP-15

X

AP-6

AP-5

AP-5

X

X

AP-9

X

AP-10

X

Best

CPCI First Hop Neighbor Second Hop Neighbor
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

59

DCA Solution Flow
AP
AP-4 AP-1 AP-7 AP-8 AP-23 AP-14 AP-13 AP-19 AP-24 AP-25 AP-16 AP-17 AP-2 AP-20 AP-21 AP-22 AP-9 AP-18 AP-3 AP-12

CM
34 60 63 67 68 71 73 75 76 77 78 81 82 83 83 84 85 87 90 91

Worst

AP-7 AP-2

AP-8
AP-5

AP-19

AP-18

AP-22 Best

CPCI First Hop Neighbor Second Hop Neighbor

The Process Begins Again with the Remaining APs on the List Randomization Is Applied for Selection of the Next CPCI

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

60

Redesigned DCA Benefits
  Faster Convergence– calculations for an RF group are much faster – can complete 6 iterations in the previous time it took for one.   More Granular – more flexible for the dynamic needs of an RF Neighborhood   System wide View – every AP’s assignment is known and managed by a centralized resource   Much better for integrating Spectrum Intelligence and makes CleanAir integration exciting.

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

61

Back to Our Use Case   Don’t like RRM’s answer – what can be done?   Change the question!

Move an AP on Either Floor Override Global for Just 1 AP and Let DCA Recalculate!

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

62

Case 4 – Transmit Power Control
  New construction   Predictive site survey done for Vocera 11.b badges   Predictive survey called for 25 – 30 foot spacing   Power at 13 dBm power (power level 3) to cover   TPC forced AP’s to power level 7   Result was coverage holes for Voice

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

63

Voice Readiness Tool Results

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

64

New Building Borders with Existing Building and AP1130 Installation

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

65

Add APs – Fill Coverage Holes without Increasing Co-Channel Interference

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

66

TPC—How It Works
  Assume an AP’s TX neighbors hear it at the following RSSI levels (listed in decreasing order; units are [dBm])
–45, –55, –67, –75, –78, –80

  For third loudest neighbor RSSI_3rd > TPC_Threshold
TPC_Threshold = –70 dBm

  TPC would recommend a Tx power decrease   Important: The RSSI_3rd >? TPC_Threshold criterion only determines if Tx decrease is recommended
Whether the actual decrease takes place depends on hysteresis The “delta” between the current and the recommended Tx Hysterisis for a TX Power increase is 3 dB Hysterisis for a TX Power decrease is 3 dB
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

67

TPC—How It Works
  There are two main TX power scenarios that can trigger an increase
There is no third neighbor – will result in maximum power TPC Equation evaluates the recommended Tx_Ideal to be in between Tx_max and Tx_current (rather than lower than TX_current)

  Power decreases take place gradually –1 power level at a time (3 dB)   TPC power increases happen immediately

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

68

TPC 6.0 MR1 Algorithm Changes
  Several changes to how power is calculated where made in the 6.0 MR-1 release   A smoothing algorithm was added that takes into account the power levels of the next neighboring AP’s and their neighbors   In situations where there is no third neighbor – the old algorithmic behavior was to default to power level 1 (no RSSI_3rd)   With these changes, if there is no third neighbor TPC looks for any neighbors heard above the TPC threshold, and interrogates those neighbors that are heard above the current TPC Threshold   An average of averages is factored against TPC recommended power   That average is used to modify the recommendation from TPC

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

69

TPC - Example
AP-6 RX-TX Neighbor List
Neighbor Ap2
AP-6

AP-1 RX-TX Neighbor List
Neighbor Power 4 5 3 4
AP-1

RSSI -45 dBm -55 dBm

RSSI -55 dBm -58 dBm -68 dBm -71 dBm

ap2 - tx ap6 - rx

Ap3 Ap4 ap5

AP-5

Neighbor
AP-3 AP-4 AP-2

RSSI -45 dBm -55 dBm -57 dBm -79 dBm

Power 5 4 5 3

Ap3 Ap6 Ap4 ap5

•  Ap1 has 2 neighbors, ap2 and ap6 •  Ap2 has three neighbors above TPC Threshold of -70 •  Ap6 has three neighbors above TPC Threshold of -70 •  Average the power settings for all 6 neighbors •  4+5+3+5+4+5=26 26/6=4 •  Use power level 4 in smoothing algorithm for final TPC recommendation of 3 for AP1
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

AP-2 RX-TX Neighbor List

70

TPC 6.0 MR1 Min/Max Power GUI Configuration
  From the controller GUI selectWireless=>802.11a/b=>RRM-TPC

Note: Ensure You Select Apply in the Upper Right Hand Corner of the Screen to Save.
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

71

Facts
  At static power level 3 – no clients on the network – average channel utilization was 30-40%!   At power level 7- Utilization was much lower at 10% – more representative of what the unloaded network should look like   Three options under current RRM –
Add more AP’s – too late Split RF group into new group Risky – live hospital borders 1 full side of the new building – separation was 6 meters Use TPC Min Max settings to raise power levels in this building Better – less risk of affecting neighboring AP’s

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

72

Use TPC Min/Max
  Set TPC Min/Max to 9 dBm and 6 dBm   Network settled at power level 5   Eliminated 1, 2 Mbps   AP’s stayed at power level 5   Channel Utilization Dropped   Voice Survey showed good coverage

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

73

TPC Min/Max Power GUI Configuration
  From the controller GUI selectWireless=>802.11a/b=>RRM-TPC

Note: Ensure You Select Apply in the Upper Right Hand Corner of the Screen to Save.
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

74

BandSelect

Access Point Assisted 5 GHz Band Selection

Challenge
Dual-Band Clients Persistently Connect to 2.4 GHz   2.4GHz may have 802.11b/g clients causing contention   2.4GHz is prone to interference
Dual-Band Client Radio 2.4/5GHz

Solution
BandSelect Directs Clients to 5 GHz Optimizing RF usage   Better usage of the higher capacity 5GHz band   Frees up 2.4 GHz for single band clients
Discovery Probes Looking for AP

Discovery Response

2.4
802.11n

5

Optimized RF Utilization by Moving 5 GHz Capable Client Out of the Congested 2.4 GHz Channels
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

75

BandSelect

Configuration – Per-SSID Override (Cont.)

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

76

The Problem
X

Beam Strength Not Directed to Client
Beam Strength

802.11a/g

802.11n

BRKEWN-3013

802.11a/g Client Connection Not Optimized, Creates Coverage Hole
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

77

The Solution
802.11a/g

Cisco Innovation: ClientLink
Up to 65% Improvement

Beam Forming

802.11n

Intelligent Beam Forming Directs Signal to Improve Performance and Coverage for 802.11a/g Devices
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

78

Higher Throughput per 11a/g Device
Up to 65% Increase in Throughput
13.6% Throughput vs. Distance 87.7% 70.4%
No Connection without ClientLink

89.5%

Test: 802.11a/g Device with 802.11n Network Source: Miercom
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

79

Higher System Capacity
Up to 27% Improvement in Channel Capacity

Channel Util of 74.2%

Channel Util of 45.2%

  Faster data transmission, less retries = more efficient use of RF channel.   Faster 11a/g transactions opens airtime for 11n devices, providing them improved experience

Test: 802.11a/g Device Measured at 16 Antenna Orientations with 802.11n Network Source: Miercom
BRKEWN-3013 © 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

80

What Is CleanAir Technology?
Detect and Classify
97 100
63 90 20 35

  CleanAir Radio ASIC   Uniquely Identify and Track Multiple Interferers   Assess Unique Impact to Wi-Fi Performance   Monitor AirQuality

Cisco CleanAir
BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

High-Resolution Interference Detection and Classification Logic Embedded into Cisco’s 802.11n Radio ASIC. Inline Operation with no CPU or Performance Impact.
Cisco Public

81

Wi-Fi and Spectrum Knowledge – Why Is Silicon Important?
  A Wi-Fi chip is a communications processor – a MODEM   It only knows
Energy that can be demodulated = Wi-Fi Energy that can not be demodulated = Noise

  Noise is complicated –
Collisions, fragments, corruption Wi-Fi that is below sensitivity threshold of the receiver

  Peaks in Wi-Fi activity can cause all of the above to occure

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

82

High Resolution Spectral Advantage
The Industry’s ONLY In-Line High-Resolution Spectrum Analyzer
Typical Wi-Fi Chipset Spectral Resolution at 5 MHz Cisco CleanAir Wi-Fi Chipset Spectral Resolution at 78 to 156 KHz

BlueTooth BlueTooth

‘Chip View Visualization’ of Microwave Oven and BlueTooth Interference

Power

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

83

Power

Microwave Oven

Microwave Oven

Benefits of CleanAir Technology
Self Healing and Optimizing

Wireless Security

Forensics for Troubleshooting

Policy Enforcement

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

84

Self Healing and Optimization
11

PERFORMANCE AIR QUALITY

6 1 RRM

Wireless LAN Controller
BRKEWN-3013

Channels 11, 6 and 1 Are Optimized for Maximum Performance and Minimum Interference
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

85

Self Healing and Optimization
11

PERFORMANCE AIR QUALITY

6 1 RRM 11

Wireless LAN Controller
BRKEWN-3013

Scanning 11, 6 and 1 Are Optimized Interference on Channel 6. Air ChannelsAvailable Channels…Quality Is Affected. RRM Is Browsing the for Maximum Performance and List of Preferred Channels to Minimum Interference Resolve Conflict…
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

6 1
86

Self Healing and Optimization
11

PERFORMANCE AIR QUALITY

6 11 1 RRM 11

Wireless LAN Controller
BRKEWN-3013

Conflict Resolved. Information Is Being Changing to Channel 11 Relayed to RRM. Conflicting Channel Is Blocked from Future Use.
© 2011 Cisco and/or its affiliates. All rights reserved. Cisco Public

X 6 1
87

Self Healing and Optimization
Interference Aware RRM Event Driven RRM Persistent Device Avoidance

Maximizes Performance by Avoiding Interference

CH 1

CH 1

CH 11

CH 1

Self Healing to Avoid Wi-Fi Degradation

Self Learning to Increase Reliability

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

88

“RF Matters”

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

89

Q&A

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

90

Links
  Cisco CleanAir solutions   Farpoint Tech Note: Evaluating Interference in Wireless LANs: Recommended Practice (PDF; 220 KB)   Farpoint Tech Note: Interference and Metro-Scale Wi-Fi Mesh Networks (PDF; 98 KB)   Farpoint Tech Note: The Effects of Interference on Video Over Wi-Fi (PDF; 100 KB)   Farpoint Tech Note: The Effects of Interference on VoFi Traffic (PDF; 88 KB)   Farpoint Tech Note: The Invisible Threat: Interference and Wireless LANs (PDF; 83 KB)   Farpoint Tech Note: The Effects of Interference on General WLAN Traffic (PDF; 88 KB)   Protecting Wi-Fi Networks from Hidden Layer 1 Security Threats (PDF; 7 MB)   RF Spectrum Policy: Future-Proof Wireless Investment Through Better Compliance   20 Myths of Wi-Fi Interference

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

91

Complete Your Online Session Evaluation
  Receive 25 Cisco Preferred Access points for each session evaluation you complete.   Give us your feedback and you could win fabulous prizes. Points are calculated on a daily basis. Winners will be notified by email after July 22nd.   Complete your session evaluation online now (open a browser through our wireless network to access our portal) or visit one of the Internet stations throughout the Convention Center.   Don’t forget to activate your Cisco Live and Networkers Virtual account for access to all session materials, communities, and on-demand and live activities throughout the year. Activate your account at any internet station or visit www.ciscolivevirtual.com.

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

92

Visit the Cisco Store for Related Titles http://theciscostores.com

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

93

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

94

Thank you.

BRKEWN-3013

© 2011 Cisco and/or its affiliates. All rights reserved.

Cisco Public

95