Page 1
1 1Vivek Shrivastava
PIE in the Sky : Online Passive Interference Estimation for Enterprise WLANs
WiNGS Labs
Vivek Shrivastava* Nokia Research Center, Palo Alto
NSDI 2011
Shravan Rayanchu, Suman Banerjee University of Wisconsin-Madison
Konstantina Papagiannaki Intel Labs, Pittsburgh
*[email protected]
Page 2
2
Enterprise WLAN setting
2Vivek Shrivastava
Wireless controller
Access Point
Clients
Internet
NSDI 2011
Page 3
3
Enterprise WLAN setting
3Vivek Shrivastava
Wireless controller
Access Point
Clients
Internet
Functionalities implemented at controller• Intrusion detection system• Interference management (channel assignment, power control)
NSDI 2011
Page 4
4
Enterprise WLAN setting
4Vivek Shrivastava
Wireless controller
Access Point
Clients
Internet
Functionalities implemented at controller• Intrusion detection system• Interference management (channel assignment, power control)
NSDI 2011
Page 5
5 5Vivek Shrivastava
Problems with wireless*
Support
“Flaky how ?”
“We will take a look.”
“The wireless is being flaky.”
“Well my connection dropped earlier and now it seems to be slow”
“Wait, now it seems fine.”User
NSDI 2011
*Slide borrowed from Cheng et. al (Jigsaw, Sigcomm ’06)
Page 6
6 6
Vivek Shrivastava
Problems with wireless
NSDI 2011 6
Hidden terminals
Page 7
7 7Vivek Shrivastava
Mismatch in data rates, slows down fast links
Problems with wireless
Rate anomaly
6 Mbps
54 Mbps
NSDI 2011
Page 8
8 8
Vivek Shrivastava
Problems with wireless
NSDI 2011 8
Increasing client density and mobility ….
Page 9
9 9
Vivek Shrivastava
Problems with wireless
NSDI 2011 9
Increasing client density and mobility….
Page 10
10
10
Vivek Shrivastava
Problems with wireless
NSDI 2011 10
…. changing interference patterns
Page 11
11
11
Vivek Shrivastava
Interference management in WLANs
NSDI 2011 11
Page 12
12
12
Vivek Shrivastava
Interference management in WLANs
Estimate Interference dynamically
Manage Interference (data scheduling, transmit power control, channel
assignment)
NSDI 2011 12
Page 13
13
13
Vivek Shrivastava
Interference management in WLANs
Manage Interference (data scheduling, transmit power control, channel
assignment)
NSDI 2011 13
Estimate Interference dynamically
Page 14
14
14Vivek Shrivastava
How to estimate interference ?
Use bandwidth tests
NSDI 2011
Page 15
15
15Vivek Shrivastava
How to estimate interference ?
Use bandwidth tests
Interferer
AP-Client pair
NSDI 2011
Page 16
16
16Vivek Shrivastava
How to estimate interference ?
NSDI 2011
1) Measure AP-Client delivery in isolation
Page 17
17
17Vivek Shrivastava
How to estimate interference ?
Isolation delivery= 0.95
NSDI 2011
1) Measure AP-Client delivery in isolation
Page 18
18
18Vivek Shrivastava
How to estimate interference ?
NSDI 2011
1) Measure AP-Client delivery in isolation
2) Activate interferer and measure delivery
Page 19
19
19Vivek Shrivastava
How to estimate interference ?
Interference delivery= 0.66
NSDI 2011
1) Measure AP-Client delivery in isolation
2) Activate interferer and measure delivery
Page 20
20
20Vivek Shrivastava
How to estimate interference ?
1) Measure AP-Client delivery in isolation
2) Activate interferer and measure delivery
NSDI 2011
Link Interference Ratio (LIR) = del Interference / del isolation
Page 21
21
21Vivek Shrivastava
How to estimate interference ?
NSDI 2011
1) Measure AP-Client delivery in isolation
2) Activate interferer and measure delivery
Link Interference Ratio (LIR) = 0.66 / 0.95 = 0.69
Page 22
22
22Vivek Shrivastava
How to estimate interference ?
NSDI 2011
1) Measure AP-Client delivery in isolation
2) Activate interferer and measure delivery
LIR0 1
Strong Weak
Page 23
23
But are bandwidth tests practical ?
•Can we use bandwidth tests in live settings
• Good accuracy –
• Network downtime required - X
• Not scalable (~ 1 hr for 20 AP-Client pair network) - X
• Not based on realistic rates and packet sizes – X
• Inefficient in dynamic scenario (client mobility) – X
23Vivek Shrivastava NSDI 2011
Page 24
24
But are bandwidth tests practical ?
•Can we use bandwidth tests in live settings
• Good accuracy –
• Network downtime required - X
• Not scalable (~ 1 hr for 20 AP-Client pair network) - X
• Not based on realistic rates and packet sizes – X
• Inefficient in dynamic scenario (client mobility) – X
24Vivek Shrivastava NSDI 2011
Can we estimate interference in a passive, real-time way ?
Page 25
25
PIE Outline
•Motivation
•Conventional bandwidth tests not sufficient
•Passive Interference Estimation (PIE)
•Polling period of PIE
•Accuracy of PIE
•Realistic trace replay with PIE
•Applications of PIE
•Summary
25Vivek Shrivastava NSDI 2011
Page 26
26
26Vivek Shrivastava
Estimating interference passively
Sniffer
Sniffer
NSDI 2011
Page 27
27
27Vivek Shrivastava
Estimating interference passively
Sniffer
Sniffer
NSDI 2011
• Sniffer could be a dedicated wireless radio• Clocks synchronized using wired backplane
Page 28
28
28Vivek Shrivastava
Estimating interference passively
Sniffer reports
NSDI 2011
Page 29
29
29Vivek Shrivastava
Estimating interference passively
Sniffer reports
Timestamp, duration, rate, success..
NSDI 2011
Page 30
30
30Vivek Shrivastava
Estimating interference passively
Hidden terminals
NSDI 2011
Page 31
31
31Vivek Shrivastava
Hidden terminals
Estimating interference passively
NSDI 2011
Page 32
32
32Vivek Shrivastava
Hidden terminals
Estimating interference passively
NSDI 2011
Page 33
33
33Vivek Shrivastava
1. Carrier sense
Hidden terminals
Estimating interference passively
NSDI 2011
Page 34
34
34Vivek Shrivastava
2. Channel free, transmit
Hidden terminals
Estimating interference passively
1) Note timestamp, rate, duration
NSDI 2011
Page 35
35
35Vivek Shrivastava
3. Collision !
Estimating interference passively
1) Note timestamp, rate, duration
2) Note if transmission is a success (ack received ?)
NSDI 2011
Page 36
36
36Vivek Shrivastava
Estimating interference passivelyTimestamp: T0Rate: 6MbpsLength: 1400 bytesSuccess: False
Timestamp: T0 + δRate: 12MbpsLength: 600 bytesSuccess: False
NSDI 2011
Page 37
37
37Vivek Shrivastava
Estimating interference passively
T06Mbps
1400 bytesFalse
T0 + δ12Mbps
600 bytesFalse
Interference estimation
NSDI 2011
Page 38
38
38Vivek Shrivastava
Estimating interference passively
T06Mbps
1400 bytesFalse
T0 + δ 12Mbps
600 bytesFalse
NSDI 2011
Page 39
39
39Vivek Shrivastava
Estimating interference passively
T06Mbps
1400 bytesFalse
T0 + δ 12Mbps
600 bytesFalse
NSDI 2011
Page 40
40
40Vivek Shrivastava
Estimating interference passively
T06Mbps
1400 bytesFalse
T0 + δ12Mbps
600 bytesFalse
&
NSDI 2011
Page 41
41
41Vivek Shrivastava
Estimating interference passively
T06Mbps
1400 bytesFalse
T0 + δ12Mbps
600 bytesFalse
&
Red & Green clients
interfere
NSDI 2011
Page 42
42
Estimating interference passively
Sniffer reports
Infer interference
Scenarios
Reception
XX
XX
Vivek Shrivastava NSDI 2011
Page 43
NSDI 201143
Estimating interference passively
Sniffer reports
Infer interference
Scenarios
Reception
XX
XX
Red and Green packets overlaps
=> both lostVivek Shrivastava
Page 44
44
Estimating interference passively
Sniffer reports
Infer interference
Scenarios
Reception
XX
XX
No overlap, no problem !
Vivek Shrivastava NSDI 2011
Page 45
Estimating interference passively
Sniffer reports
Infer interference
Scenarios
Reception
XX
XX Both way
hidden terminals
Vivek Shrivastava 45NSDI 2011
Page 46
46
Estimating interference passively
Sniffer reports
Infer interference
Scenarios
Reception
X X
Vivek Shrivastava NSDI 2011
Page 47
47
Estimating interference passively
Sniffer reports
Infer interference
Scenarios
Reception
X X
Red and Green packets overlaps=> Green is lostVivek Shrivastava
Page 48
48
Estimating interference passively
Sniffer reports
Infer interference
Scenarios
Reception
X X One way hidden
terminalsVivek Shrivastava NSDI 2011
Page 49
49
Computing interference measure in PIE
•Compute Isolation loss rate
•Fraction of non-overlapping packets lost
•Compute Interference loss rate
•Fraction of overlapping packets lost
•Interference measure (LIR):
(1 – Interference loss) / (1 – Isolation loss )
49Vivek Shrivastava NSDI 2011
Page 50
50
How quickly can PIE converge ?
•Time taken by PIE to converge depends on two key properties
• Periodicity with which sniffer reports are collected by the controller
•Traffic patterns for the links which dictate the number of interference events captured in a time interval
50Vivek Shrivastava NSDI 2011
Page 51
51
How quickly can PIE converge ?
•Time taken by PIE to converge depends on two key properties
• Periodicity with which sniffer reports are collected by the controller
•What is the minimum polling period ?
•Traffic patterns for the links which dictate the number of interference events captured in a time interval
•How much time does PIE take under realistic access patterns ?
51Vivek Shrivastava NSDI 2011
Page 52
52
PIE Outline
•Motivation
•Conventional bandwidth tests not sufficient
•Passive Interference Estimation (PIE)
•Polling period of PIE
•Accuracy of PIE
•Realistic trace replay with PIE
•Applications of PIE
•Summary
52Vivek Shrivastava NSDI 2011
Page 53
53
53Vivek Shrivastava
What is the minimum polling period ?
TimeI1 I2time interval
P P P
NSDI 2011
Page 54
54
54Vivek Shrivastava
TimeI1 I2time interval
P P P
R(I1) R(I1)
NSDI 2011
What is the minimum polling period ?
Page 55
55
55Vivek Shrivastava
What is the minimum polling period ?
TimeI1 I2time interval
P P P
LIR (I1) R(I1) R(I1)
NSDI 2011
Page 56
56
56Vivek Shrivastava
What is the minimum polling period ?
TimeI1 I2time interval
P P P
R(I2) R(I2)
LIR (I1)
NSDI 2011
Page 57
57
57Vivek Shrivastava
What is the minimum polling period ?
TimeI1 I2time interval
P P P
LIR (I2).
R(I2) R(I2)
LIR (I1)
NSDI 2011
Page 58
58
What is the minimum polling period ?
58Vivek Shrivastava
LIR
Polling period (ms)Stability of interference
measure for saturated trafficNSDI 2011
Page 59
59
What is the minimum polling period ?
59Vivek Shrivastava
Measure stabilizes after ~85 ms (at least 20 overlap samples)LI
R
Polling period (ms)Stability of interference
measure per polling periodNSDI 2011
Page 60
60
What is the minimum polling period ?
60Vivek Shrivastava
LIR
Polling period (ms)Stability of interference
measure per polling periodNSDI 2011
We use a polling period of 100ms
Page 61
61
PIE Outline
•Motivation
•Conventional bandwidth tests not sufficient
•Passive Interference Estimation (PIE)
•Polling period of PIE
•Accuracy of PIE
•Realistic trace replay with PIE
•Applications of PIE
•Summary
61Vivek Shrivastava NSDI 2011
Page 62
62
How accurate is PIE ?
62Vivek Shrivastava
% o
f lin
k-in
terf
ere
r p
air
s
Mean Error in LIR estimation
Page 63
63
How accurate is PIE ?
63Vivek Shrivastava
% o
f lin
k-in
terf
ere
r p
air
s
Mean Error in LIR estimation
Page 64
64
How accurate is PIE ?
64Vivek Shrivastava
% o
f lin
k-in
terf
ere
r p
air
s
Mean Error in LIR estimation
95% of link-interferer pairs, LIR computed by PIE is within +/- 0.1 of the value reported by BW test
Page 65
65
PIE Outline
•Motivation
•Conventional bandwidth tests not sufficient
•Passive Interference Estimation (PIE)
•Polling period of PIE
•Accuracy of PIE
•Realistic trace replay with PIE
•Applications of PIE
•Summary
65Vivek Shrivastava NSDI 2011
Page 66
66
66Vivek Shrivastava
PIE with realistic access patterns
NSDI 2011
Page 67
67
67Vivek Shrivastava NSDI 2011
PIE with realistic access patterns
• Evaluate PIE using realist traffic patterns on a 15 node topology (7 AP – 8 laptops)
• Each client laptop replays the traffic patterns of an actual client from a real wireless trace
• Three activity periods: heavy (> 40 % medium busy), medium (40 – 20% busy), light (< 20% busy)
Page 68
68
PIE with realistic access patterns
68Vivek Shrivastava
Tim
e to
est
imat
e (m
s)
NSDI 2011
Traffic period
Heavy Medium Light 0
200
400
600
800
1000
1200
Page 69
69
PIE with realistic access patterns
69Vivek Shrivastava
Tim
e to
est
imat
e (m
s)
NSDI 2011
Traffic period
Heavy Medium Light 0
200
400
600
800
1000
1200
• Convergence is faster for higher client activity
• Even for light activity, median time of estimate LIR is less than 650 ms
Page 70
70
PIE Outline
•Motivation
•Conventional bandwidth tests not sufficient
•Passive Interference Estimation (PIE)
•Polling period of PIE
•Accuracy of PIE
•Realistic trace replay with PIE
•Applications of PIE
•Summary
70Vivek Shrivastava NSDI 2011
Page 71
71
71Vivek Shrivastava
What is the impact on WLAN applications?
NSDI 2011
AP-Client pairs
Page 72
72
72Vivek Shrivastava
What is the impact on WLAN applications?
NSDI 2011
AP-Client pairs
Evaluate usefulness of PIE for an interference mitigation mechanism (data scheduling using CENTAUR – Mobicom ‘09)
Page 73
73
73Vivek Shrivastava
What is the impact on WLAN applications?
NSDI 2011
AP-Client pairs
1. Estimate interference using PIE
Page 74
74
74Vivek Shrivastava
What is the impact on WLAN applications?
NSDI 2011
AP-Client pairs
1. Estimate interference using PIE2. Input estimate to a centralized data scheduler
Page 75
75
75Vivek Shrivastava
What is the impact on WLAN applications?
NSDI 2011
AP-Client pairs
1. Estimate interference using PIE2. Input estimate to a centralized data scheduler3. Evaluate performance under dynamic scenarios
Page 76
76
What is the impact on end users ?
76Vivek Shrivastava
Sys
tem
T
hrou
ghpu
t (M
bps)
Static scenario
Distributed
Cent. Sched (BW)
Cent. Sched (PIE)
0
2
4
6
8
10
12
14
Static
NSDI 2011
Page 77
Distributed
Cent. Sched (BW)
Cent. Sched (PIE)
0
2
4
6
8
10
12
14
Static
77
What is the impact on end users ?
77Vivek Shrivastava
Sys
tem
T
hrou
ghpu
t (M
bps)
Static scenario
NSDI 2011
Page 78
Distributed
Cent. Sched (BW)
Cent. Sched (PIE)
0
2
4
6
8
10
12
14
Static
78
What is the impact on end users ?
78Vivek Shrivastava
Sys
tem
T
hrou
ghpu
t (M
bps)
Static scenario
Static scenarios, PIE is comparable to BW test
NSDI 2011
Page 79
79
What is the impact on end users ?
79Vivek Shrivastava
Sys
tem
T
hrou
ghpu
t (M
bps)
Mobile scenario
Distributed
Cent. Sched. (BW)
Cent. Sched. (PIE)
0
2
4
6
8
10
12
14
Series2
NSDI 2011
Page 80
Distributed
Cent. Sched. (BW)
Cent. Sched. (PIE)
0
2
4
6
8
10
12
14
Series2
80
What is the impact on end users ?
80Vivek Shrivastava
Sys
tem
T
hrou
ghpu
t (M
bps)
Mobile scenario
NSDI 2011
Page 81
Distributed
Cent. Sched. (BW)
Cent. Sched. (PIE)
0
2
4
6
8
10
12
14
Series2
81
What is the impact on end users ?
81Vivek Shrivastava
Sys
tem
T
hrou
ghpu
t (M
bps)
Mobile scenario
Mobility scenarios, PIE outperforms BW test
NSDI 2011
Page 82
82
What is the impact on end users ?
•PIE can also be used to monitor production systems (like Jigsaw)
•We monitored two production WLANs
•Use testbed nodes in proximity of production APs as sniffers
•Identify hidden terminals and rate anomaly problems
82Vivek Shrivastava NSDI 2011
Page 83
83
What is the impact on end users ?
83Vivek Shrivastava NSDI 2011
WLANs
WLAN1
WLAN2
Hidden terminal cases (LIR < 0.7)
Rate anomaly cases (Ratio of rates < 0.2)
8%
11%
21%
22%
Page 84
84
What is the impact on end users ?
84Vivek Shrivastava NSDI 2011
WLANs
WLAN1
WLAN2
Hidden terminal cases (LIR < 0.7)
Rate anomaly cases (Ratio of rates < 0.2)
8%
11%
21%
22%
•Hidden terminals are rare, but can become pain points for clients
•Rate anomaly is more frequent, but do not cause drastic performance issues
Page 85
85
PIE Outline
•Motivation
•Conventional bandwidth tests not sufficient
•Passive Interference Estimation (PIE)
•Polling period of PIE
•Accuracy of PIE
•Realistic trace replay with PIE
•Applications of PIE
•Summary
85Vivek Shrivastava NSDI 2011
Page 86
86
Related Work
•PIE leverages techniques from Jigsaw, WIT (Sigcomm 2006) and builds on their ideas
• Focus of Jigsaw, WIT was to understand interference, ours is to compute it in real-time
• CMAP also infers interference to harness exposed terminals, but requires physical layer change
•Active techniques like Microprobing (CoNext 2008) still require downtime and do not use realistic traffic
86Vivek Shrivastava NSDI 2011
Page 87
87
PIE Limitations
•Does not handle non-WiFi interferer like microwaves.
•Can miss external interferers if none of the enterprise APs can listen to the interferer
• May miss client conflicts, can use client participation in PIE to enhance the system
•Interference detection techniques at the physical layer may be more accurate in some scenarios where diversity is too low for PIE to function
87Vivek Shrivastava NSDI 2011
Page 88
88
PIE Summary
•Online interference estimation important for interference mitigation
•BW test incurs high overhead, requires downtime
•PIE is a passive mechanism, generates interference estimates in real time
•Leverages centralized infrastructure to collect real time reports from APs
•Non-intrusive with good accuracy
88Vivek Shrivastava NSDI 2011
Page 89
Vivek Shrivastava 89
Thank you !
[email protected] www.cs.wisc.edu/~viveks
NSDI 2011
Page 90
90
Is there sufficient diversity ?
90Vivek Shrivastava
% Overlap in transmit time
% o
f tr
ansm
it
pair
s
NSDI 2011
Page 91
91
Is there sufficient diversity ?
91Vivek Shrivastava
% Overlap in transmit time
% o
f tr
ansm
it
pair
s
Overlap in transmit times for transmitter pairs in real WLAN traces
NSDI 2011
Page 92
92
Is there sufficient diversity ?
92Vivek Shrivastava
% Overlap in transmit time
% o
f tr
ansm
it
pair
s
80% of transmit pairs have less then 10%
overlap
Overlap in transmit times for transmitter pairs in real WLAN traces
NSDI 2011
Page 93
93
How quick is PIE ?
93Vivek Shrivastava
Convergence time of PIE under varying loads
Traffic load on link and interferer
Con
verg
en
ce
tim
e
NSDI 2011
Page 94
94
How quick is PIE ?
94Vivek Shrivastava
Convergence time of PIE under varying loads
Traffic load on link and interferer
Con
verg
en
ce
tim
e
NSDI 2011
Page 95
95
How quick is PIE ?
95Vivek Shrivastava
Convergence time of PIE under varying loads
Traffic load on link and interferer
Con
verg
en
ce
tim
e
Converges within 650 ms for light traffic loads as well
NSDI 2011
Page 96
96
Synchronization error in PIE ?
96Vivek Shrivastava
Synchronization error using TSF beacon synchronizationNSDI 2011
Page 97
97
97Vivek Shrivastava
How accurate is PIE ?
InterfererAP-Client pair
Non Interferer
1. Activate link, interferer and non-interferer
NSDI 2011
Page 98
98
98Vivek Shrivastava
How accurate is PIE ?
InterfererAP-Client pair
Non Interferer
1. Activate link, interferer and non-interferer2. Measure LIR for interferer and non-interferer
NSDI 2011
Page 99
99
99Vivek Shrivastava
What is the impact on WLAN applications?
NSDI 2011
AP-Client pairs
Page 100
100
100Vivek Shrivastava
What is the impact on end users ?AP-Client
pairs
NSDI 2011
Page 101
101
101Vivek Shrivastava
What is the impact on end users ?AP-Client
pairs
Evaluate PIE on a 7 AP - 8 Client topology
NSDI 2011
Page 102
102
102Vivek Shrivastava
What is the impact on end users ?AP-Client
pairs
Evaluate usefulness of PIE for an interference mitigation mechanism
NSDI 2011
Page 103
103
103Vivek Shrivastava
What is the impact on end users ?AP-Client
pairs
1. Estimate interference using PIE
NSDI 2011
Page 104
104
104Vivek Shrivastava
What is the impact on end users ?AP-Client
pairs
1. Estimate interference using PIE2. Input estimate to a centralized data scheduler
NSDI 2011
Page 105
105
105Vivek Shrivastava
What is the impact on end users ?AP-Client
pairs
1. Estimate interference using PIE2. Input estimate to a centralized data scheduler3. Evaluate performance under dynamic scenarios
NSDI 2011
Page 106
106
106Vivek Shrivastava
Carrier sensing (conservative)
Problems with wireless
Exposed terminals
NSDI 2011
Page 107
107
107
Vivek Shrivastava
Interference management in WLANs
Manage Interference (data scheduling, transmit power control, channel
assignment)
NSDI 2011 107
Estimate Interference
Page 108
108
108Vivek Shrivastava
Key Challenge
Q. Is there a way to dynamically identify the precise set of nodes in an enterprise WLAN that interfere with each other and the degree
to which they do so ?
NSDI 2011
Page 109
109
PIE Outline
•Motivation
•Conventional bandwidth tests not sufficient
•Passive Interference Estimation (PIE)
•Accuracy of PIE
•Convergence of PIE
•Agility of PIE
•Applications of PIE (data scheduling)
•Summary
109Vivek Shrivastava NSDI 2011
Page 110
110
110Vivek Shrivastava
How agile is PIE ?
Client moves from AP towards interferer
Interferer
AP-Client pair
NSDI 2011
Page 111
111
111Vivek Shrivastava
How agile is PIE ?
Client moves from AP towards interferer
Interferer
AP-Client pair
NSDI 2011
Page 112
112
112Vivek Shrivastava
How agile is PIE ?
Client moves from AP towards interferer
Interferer
AP-Client pair
NSDI 2011
Page 113
113
How agile is PIE ?
113Vivek Shrivastava
InterfererAP
Tracking client mobility.
SNR
SN
R (
dB)
NSDI 2011
Page 114
114
How agile is PIE ?
114Vivek Shrivastava
InterfererAP
Tracking client mobility.
SNR
Tput
SN
R (
dB)
Tpu
t (M
bps)
NSDI 2011
Page 115
115
How agile is PIE ?
115Vivek Shrivastava
Tracking client mobility.
SNR
Tput
LIRS
NR
(dB
)T
put (
Mbp
s)LI
R (
PIE
)
NSDI 2011
Page 116
116
116Vivek Shrivastava
% Overlap in transmit time
LIR
NSDI 2011
Can PIE classify interferers accurately ?
Page 117
117
Can PIE classify interferers accurately ?
117Vivek Shrivastava
% Overlap in transmit time
LIR
NSDI 2011
Page 118
118
Can PIE classify interferers accurately ?
118Vivek Shrivastava
% Overlap in transmit time
LIR
When overlap is less than 80%, PIE is able to classify accurately
NSDI 2011
Page 119
119
Convergence of PIE•Test stability of interference measure with different
polling periods
•Polling period of 85 ms is sufficient for stable interference measure in presence of saturated traffic
•Test stability of interference measure with different traffic rates
•Higher convergence times for low traffic rates, as number of events per polling period is low
•Still converges within 600ms for very lightly loaded links (when load is 2-3% of link capacity)
119Vivek Shrivastava NSDI 2011
Page 120
120
Estimating carrier sense passively
Sniffer reports
Infer carrier sense
Vivek Shrivastava NSDI 2011
Page 121
121
Estimating carrier sense passively
Sniffer reports
Infer carrier sense
Vivek Shrivastava NSDI 2011
Determine CS by observing direction of packet overlaps for contending
transmitters
Page 122
122
Estimating carrier sense passively
Sniffer reports
Infer carrier sense
Scenarios
Vivek Shrivastava NSDI 2011
Observe packet overlaps in both directions
Page 123
123
Estimating carrier sense passively
Sniffer reports
Infer carrier sense
Scenarios
Vivek Shrivastava NSDI 2011
Both Red and Green APs do not sense each
otherObserve packet overlaps
in both directions
Page 124
124
Estimating carrier sense passively
Sniffer reports
Infer carrier sense
Scenarios
Vivek Shrivastava NSDI 2011
Red senses Green, but
not vice versa
Observe packet overlaps in only one direction
Page 125
125
Estimating carrier sense passively
Sniffer reports
Infer carrier sense
Scenarios
Vivek Shrivastava NSDI 2011
Green senses Red, but not vice
versa
Observe packet overlaps in only one direction
Page 126
126
Estimating carrier sense passively
Sniffer reports
Infer carrier sense
Scenarios
Vivek Shrivastava NSDI 2011
No contending packets overlap
Green and Red sense each other
Page 127
127
Computing carrier sense measure in PIE
•Compute overlap probability for contending packets ( Poverlap )
•If (1 – Poverlap ) > δt , report no carrier sensing
•else if Poverlap > δt
•If Pone-way > δt, report one-way CS
•else report mutual carrier sensing
127Vivek Shrivastava NSDI 2011
Page 128
128
128Vivek Shrivastava
How accurate is PIE ?
Interferer
AP-Client pair
NSDI 2011
Page 129
129
129Vivek Shrivastava
How accurate is PIE ?
Interferer
AP-Client pair
1. Activate both link and interferer
NSDI 2011
Page 130
130
130Vivek Shrivastava
How accurate is PIE ?
Interferer
AP-Client pair
1. Activate both link and interferer2. Measure LIR using PIE and BW test
Page 131
131
How accurate is PIE with multiple interferers ?
Scenarios
Reception
X X
Vivek Shrivastava
Red and Green packets overlap=> Green is lost
One way hidden
terminals
NSDI 2011
Page 132
132
How accurate is PIE with multiple interferers ?
Scenarios
Reception
Vivek Shrivastava
Blue and Green packets overlap=> None is lost
No interference
NSDI 2011
Page 133
133
Is there sufficient diversity ?
133Vivek Shrivastava
% Overlap in transmit time
% o
f tr
ansm
it
pair
s
NSDI 2011
Page 134
NSDI 2011134
Is there sufficient diversity ?
134Vivek Shrivastava
% Overlap in transmit time
% o
f tr
ansm
it
pair
s
Overlap in transmit times for transmitter pairs in real WLAN traces
Page 135
NSDI 2011135
Is there sufficient diversity ?
135Vivek Shrivastava
% Overlap in transmit time
% o
f tr
ansm
it
pair
s
80% of transmit pairs have less then 5%
overlap
Overlap in transmit times for transmitter pairs in real WLAN traces
Page 136
136
What about multiple interferers?
X
Vivek Shrivastava
X
Ambiguous of loss source
(Red/Blue)
NSDI 2011
Page 137
137
What about multiple interferers?
X
Vivek Shrivastava
X
Overlap with Blue
=>No Loss
NSDI 2011
Page 138
138
What about multiple interferers?
X
Vivek Shrivastava
X
Overlap with Red => Loss
NSDI 2011
Page 139
139
What about multiple interferers?
X
Vivek Shrivastava
X
Isolation =>No Loss
NSDI 2011
Page 140
140
What about multiple interferers?
X
Vivek Shrivastava
X
Isolation =>No Loss
Overlap with Blue
=>No Loss
NSDI 2011
Page 141
141
What about multiple interferers?
X
Vivek Shrivastava
X
Isolation =>No Loss
Overlap with Blue
=>No Loss
Overlap with Red => Loss
NSDI 2011
Page 142
142
What about multiple interferers?
X
Vivek Shrivastava
X
PIE needs transmission diversity to identify interferers accurately
NSDI 2011
Page 143
143
PIE with realistic access patterns
143Vivek Shrivastava
% o
f lin
k -
inte
rfer
er
pairs
NSDI 2011
Convergence time (ms)
Page 144
144
PIE with realistic access patterns
144Vivek Shrivastava
% o
f lin
k -
inte
rfer
er
pairs
NSDI 2011
Convergence time (ms)
Page 145
145
PIE with realistic access patterns
145Vivek Shrivastava
% o
f lin
k -
inte
rfer
er
pairs
NSDI 2011
Convergence time (ms)
• Convergence is faster for higher client activity
Page 146
146
PIE with realistic access patterns
146Vivek Shrivastava
% o
f lin
k -
inte
rfer
er
pairs
NSDI 2011
Convergence time (ms)• Convergence is faster for higher client activity
• Even for light client activity, 90% of link-interferer pairs converge within ~ 10 seconds
Page 147
147
Synchronization error
147Vivek Shrivastava
Synchronization error for a 19 node topology
NSDI 2011
CD
F
Synchronization error (microsec)
Page 148
148
Synchronization error
148Vivek Shrivastava
Synchronization error as a function of beacon interval
NSDI 2011
Clo
ck s
kew
(u
s)
Time (ms)