Post on 16-Jan-2016
description
tseng:1
Power Saving and Power Managementin WiFi and Bluetooth Networks
Prof. Yu-Chee Tseng
Dept. of Comp. Sci. & Infor. Eng.
National Chiao-Tung University
( 交通大學 資訊工程系 曾煜棋 )
yctseng: 2
Outline Power control:
S.-L. Wu, Y.-C. Tseng, and J.-P. Sheu, "Intelligent Medium Access for Mobile Ad Hoc Networks with Busy Tones and Power Control", IEEE Journal on Selected Areas in Communications, 18(9):1647-1657, Sep. 2000.
Power management: Y.-C. Tseng, C.-S. Hsu, and T.-Y. Hsieh, "Power-Saving Pr
otocols for IEEE 802.11-Based Multi-Hop Ad Hoc Networks", Computer Networks, Elsevier Science Pub., Vol. 43, No. 3, Oct. 2003, pp. 317-337.
WiFi vs Bluetooth: T.-Y. Lin and Y.-C. Tseng, "An Adaptive Sniff Scheduling S
cheme for Power Saving in Bluetooth", IEEE Wireless Communications, Vol. 9, No. 6, Dec. 2002, pp. 92-103.
tseng:3
Introduction: Basic Concept
4
Introduction Battery is a limited resource in any
portable device. becoming a very hot topic is wireless
communication
Power-related issues: PHY: transmission power control MAC: power mode management Network Layer: power-aware routing
5
Transmission Power Control tuning transmission energy for higher
channel reuse example:
A is sending to B (based on IEEE 802.11) Can (C, D) and (E, F) join?
No! Yes!
6
Power Mode Management doze mode vs. active mode example:
A is sending to B (based on 802.11) Does C need to stay awake?
7
Power-Aware Routing routing in an ad hoc network with
energy-saving in mind Example: in an ad hoc network
+
–
+
–
+
–
+
–
+
–
+
–
SRC
N1 N2
DEST
N4N3
wu-ICCCN99:8
WirelessNetTseng
Intelligent Medium Access for Mobile Ad Hoc Networks with Busy Tones and Power
Control
S.-L. Wu, Y.-C. Tseng, and J.-P. Sheu,
IEEE J. of Selected Areas on Communications (JSAC)
wu-ICCCN99:9
WirelessNetTseng
Abstract
A New MAC Protocol based on RTS/CTS with Busy Tones with Power Control
wu-ICCCN99:10
WirelessNetTseng
Power Control Use an appropriate power level to transmit packets.
to increase the possibility of channel reuse to increase channel utilization
Example: (a) without power control:
the transmissions from C to D and from E to F are prohibited.
(b) with power control:all these can coexist.
F
E
DC
BA
F
E
DC
BA
(b )(a )
wu-ICCCN99:11
WirelessNetTseng
How to Tune Power Levels
Assumptions: A mobile host can choose on what power level to transmit a packet. On receiving a packet, the physical layer can offer the MAC layer
the power level on which the packet was received. Suppose Pt and Pr are the power levels a packet is sent and received, re
spectively.
= carrier wavelength n = path loss coefficient (typically 2 ~ 6) d = distance between sender and receiver gt and gr: antenna gains at the sender and receiver sides, respectivel
y
rtn
tr ggd
PP )4
(
wu-ICCCN99:12
WirelessNetTseng
Note: during a short period, the values of n and d can be treated as a constant. This makes power control possible.
Let Pmin be the minimum power level to decode a packet.
Suppose X sends an RTS to Y with power Pt.
If Y wants to reply a CTS to X with a power level PCTS, such that X receives the packet at the smallest power level Pmin, then we have:
Dividing the above formulas, we have:
rtn
tr ggd
PP )4
(
rtn
CTS ggd
PP )4
(min
t
CTS
r P
P
P
Pmin
wu-ICCCN99:13
WirelessNetTseng
General Rules in This Paper
Busy Tone (BT) Senders should send BTt, but gauge any BTr. Receivers should send BTr, but gauge any BTt.
General Rules: Data packet and BTt: transmitted with power control. CTS and BTr: transmitted at the normal (largest) power. RTS: at a power level based on how strong the BTr are aroun
d the requesting host. Channel Model:
BTt BTr
controlchannel
datachannel
frequency
wu-ICCCN99:14
WirelessNetTseng
Illustrative Example (I)
A is sending to B. A’s data packet and BTt at t
he minimal level (yellow circle).
B’s BTr at the largest level (white circle).
C intends to send to D. C hears no BTr. D hears not BTt. So the transmission can be
granted (pink circle).
A
DC
B
wu-ICCCN99:15
WirelessNetTseng
Illustrative Example (II)
Now we moe C into A’s circle. A is sending to B.
A’s data packet and BTt at the minimal level (yellow circle).
B’s BTr at the largest level (white circle).
C intends to send to D. C hears no BTr. D hears no BTt. So the transmission can be
granted (pink circle).
A
DC
B
wu-ICCCN99:16
WirelessNetTseng
Illustrative Example (III)
Next we move D into A’s circle. A is sending to B.
A’s data packet and BTt at the minimal level (yellow circle).
B’s BTr at the largest level (white circle).
C intends to send to D. C hears no BTr. D hears A’s BTt. So the transmission can NO
T be granted (pink circle).
A
DC
B
wu-ICCCN99:17
WirelessNetTseng
Illustrative Example (IV)
A is sending to B. A’s data packet and BTt at t
he minimal level (yellow circle).
B’s BTr at the largest level (white circle).
C intends to send to D. C hears A’s BTt and B’s BT
r. D hears no BTt. The transmission can be gra
nted if C controls its transmission power (pink circle).
A
D
C
B
wu-ICCCN99:18
WirelessNetTseng
Illustrative Example (V)
A is sending to B. A’s data packet and BTt at t
he minimal level (yellow circle).
B’s BTr at the largest level (white circle).
C intends to send to D. C hears A’s BTt and B’s BT
r. D hears no BTt. The transmission can be gra
nted if C controls its transmission power (pink circle).
A
DC
B
wu-ICCCN99:19
WirelessNetTseng
Many Transmission Pairs with Power Control and Busy Tones
A
F
ED
C
B
BTt and DATA: yellow circlesBTr: white circles
wu-ICCCN99:20
WirelessNetTseng
The Protocol
Pmax: the maximum transmission power
Pmin: the minimum power to distinguish a signal from a noise
Pnoise: the maximum power at which an antenna will regard a signal as a noise Pmin - Pnoise should be a very small value
Basic “Power” Rules: Data packet and BTt: transmitted with power control. CTS and BTr: transmitted at the largest power Pmax.
RTS: at a power level based on how strong the BTr are around the requesting host.
wu-ICCCN99:21
WirelessNetTseng
Detailed Protocol
On a host X intending to send a RTS to Y, X senses any receive busy tone BTr around it X sends a RTS on the control channel at power level Px:
If there is no BTr, let Px = Pmax.
O/w, let Pr be the power level of BTr that has the highest power among all heard BTr’s.
The RTS should not go beyond the nearest host that is currently receiving a data packet.
Pmax is used because BTr is always transmitted at the maximal power.
r
noixex P
PPP max
wu-ICCCN99:22
WirelessNetTseng
On Y receiving X’s RTS, Y senses any transmit busy tone BTt around it.
If there is any BTt, then Y ignores this RTS.O/w, Y does the following:
reply with a CTS at the maximum power Pmax turn on its receive busy tone BTr at the maximum power Pmax
On X receiving Y’s CTS, X transmits its data packet at power Px.
X turns on its transmit busy tone BTt at power Px.
Pr is the power level at which X receives Y’s CTS. Px is the minimal possible power level to ensure that Y can correctly receive the data packet.
rx P
PPP maxmin
wu-ICCCN99:23
WirelessNetTseng
Many Transmission Pairs with Power Control and Busy Tones
F
ED
C
G H
BTr
BTt
A B
RTS
CTS
wu-ICCCN99:24
WirelessNetTseng
Analysis
Scenario: A is currently sending to B. Another pair, C and D, is intending to communicate.
Goal: We want to find out the probability that C can send to D.
Through complicated calculus, we find that …
wu-ICCCN99:25
WirelessNetTseng
When BC < rmax
INTC(Ra, Rb, AB) = the intersection of the circles centered at a and b Ra = radius of the circle centered at a Rb = radius of the circle centered at b AB = distance of a and b
The probability that C can send to D when A is sending to B:
i.e., the coverage of Rc excluding the coverage of Ra Fig. 6
wu-ICCCN99:26
WirelessNetTseng
wu-ICCCN99:27
WirelessNetTseng
cont...
Integrating over = 0 .. 2, and then over CB = 0 .. rmax
Integrating over AB = 0 .. rmax, we have the final result
On the contrary, the DBTMA has probability of 0.
wu-ICCCN99:28
WirelessNetTseng
When rmax < BC < 3rmax
Main difference: C’s RTS will be sent with max. power.
The probability that C can send to D when A is sending to B:
See Fig. 7:At point C1, node C can always send.At point C2, node C can’t send if D is in A’s range.
wu-ICCCN99:29
WirelessNetTseng
wu-ICCCN99:30
WirelessNetTseng
cont...
Integrating over = 0 .. 2, and then over CB = rmax..3rmax
Integrating over AB = 0 .. rmax, we have the final result
wu-ICCCN99:31
WirelessNetTseng
cont.
On the contrary, the DBTMA has a success probability of
X
change to rmax
wu-ICCCN99:32
WirelessNetTseng
Discrete Power Control
The levels of power provided by hardware may not be infinitely tunable. We may have a discrete number of power levels.
Theorem: Given a fixed integer k, evenly spreading the k power levels
will be the best choice. I.e., (1/k)*Pmax, (2/k)*Pmax, (3/k)*Pmax, …, (k/k)*Pmax.
wu-ICCCN99:33
WirelessNetTseng
Simulation Parameters
Simulation parameters physical area = 8km 8km max transmission distance (rmax) = 0.5 or 1.0 km number of mobile hosts = 600 Speed of mobile hosts 0 or 125 km/hr. length of control packet = 100 bits link speed = 1 Mbps transmission bit error rate = 10-5/bit
wu-ICCCN99:34
WirelessNetTseng
Simulation Results: Channel utilization vs. traffic load
(a) rmax = 0.5 km (b) rmax = 1.0 km
0
2 0
4 0
6 0
8 0
10 0
0 20 0 400 600 8 00 10 00 1200
L oa d (pa ck e t/m s)(a )
Cha
nnel
Util
itiza
tion
O ursD B TM A
0
5
1 0
1 5
2 0
2 5
3 0
0 10 0 200 300 4 00 500 600 700 800
L o ad (p ac k e t/m s)(b )
Cha
nnel
Util
itiza
tion
O urs
D B T M A
wu-ICCCN99:35
WirelessNetTseng
Channel utilization vs. data packet length at various traffic loads
20
25
30
35
40
45
1 2 4 8
Data Packet Length (Kbits)
Cha
nnel
Util
itiza
tion
Load=100 Kbit/msLoad=200 Kbit/msLoad=400 Kbit/msLoad=800 Kbit/ms
wu-ICCCN99:36
WirelessNetTseng
Channel Utilization vs. Number of Power Levels
rmax = 1 km; arrival rate = 200 or 400 packets/ms; packet length = 1 or 2 Kbits So 4 to 6 levels will be sufficient.
0
5
10
15
20
25
0 2 4 6 8 10 12 14 16 18 20
Power Level
Cha
nnel
Util
itiza
tion(
data
pac
ket
time/
ms) 27.34
wu-ICCCN99:37
WirelessNetTseng
Channel Utilization vs. Traffic Load
mobility = 0 km/hr and 125 km/hr The transmission distance rmax = 1.0 km
0
5
10
15
20
25
30
35
40
45
50
55
0 100 200 300 400 500 600 700 800
Load (packet/ms)
Cha
nnel
Uti
liti
zatio
n
packet length=1 Kbits, speed= 0 km/hourpacket length=1 Kbits, speed=125 km/hourpacket length=4 Kbits, speed= 0 km/hourpacket length=4 Kbits, speed=125 km/hour
wu-ICCCN99:38
WirelessNetTseng
Short Conclusion
a new MAC protocol power control on top of RTS/CTS and busy tones
Channel utilization can be significantly increased because the severity of signal overlapping is reduced.
39
Power Mode Management in IEEE 802.11
Y.-C. Tseng, C.-S. Hsu, and T.-Y. Hsieh, "Power-Saving Protocols for IEEE 802.11-Based Multi-Hop Ad Hoc Networks", Computer Networks, Elsevier Science Pub., Vol. 43, No. 3, Oct. 2003, pp. 317-337 (also in INFOCOM).
40
Power Consumption IEEE 802.11 power model
transmit: 1400 mW receive: 1000 mW idle: 830 mW sleep: 130 mW
41
Power Mode Management Power modes in IEEE 802.11
PS and ACTIVE
Problem Spectrum: infrastructure ad hoc network (MANET)
single-hopmulti-hop ad hoc networks
42
Infrastructure Mode two power modes: active and power-
saving (PS)
43
Ad Hoc Mode (Single-Hop) PS hosts also wake up periodically.
interval = ATIM (Ad hoc) window
Beacon Interval Beacon Interval
ATIM Window
ATIM Window
Host A
Host B
Beacon
BTA=2, BTB=5
power saving state
power saving state
Beacon
ATIM
ACK
active state
data frame
ACK
44
Problem Statement(Multi-Hop MANET)
Clock Synchronization: a difficult job due to communication delays
and mobility Neighbor Discovery:
by inhibiting other's beacons, hosts may not be aware of others’ existence
Network Partitioning: with unsynchronized ATIM windows, hosts
with different wakeup times may become partitioned networks
45
Network-Partitioning Example
Host A
Host B
A
B
C
D
E
F
D
E
F
Host C
Host D
Host E
Host F
╳
╳
ATIM window
╳
╳
Network Partition
46
What Do We Need? PS protocols for multi-hop ad hoc
networks Fully distributed No need of clock synchronization (i.e.,
asynchronous PS) Always able to go to sleep mode, if
desired
47
Features of Our Design Guaranteed Overlapping Awake
Intervals: two PS hosts’ wake-up patterns always
overlap no matter how much time their clocks drift
Wake-up Prediction: with beacons, derive other PS host's
wake-up pattern based on their time difference
48
Beacon Int. (BI)
Act. Win. (AW)
Structure of a Beacon Interval
BI: beacon interval (to send beacons) AW: active window
BW: beacon window MW: MTIM window (for receiving MTIM) listening period: to monitor the
environment
BW MW listening BW MW listening
49
Three Protocols Based on the above structure, we
propose three protocols Dominating-Awake-Interval Periodical-Fully-Awake-Interval Quorum-Based
50
P1: Dominating-Awake-Interval intuition: impose a PS host to stay
awake sufficiently long “dominating-awake” property
BWBIAW 2/
Host A
Host B
Beacon Interval
Beacon Interval
Beacon Interval
Beacon Interval
╳ ╳
51
Problem: only dectectable in ONE direction
Adjustment: odd beacon interval:
Active Window = BW + MW + listening
even beacon interval: Active Window = listening + MW + BW
Host A
Host B
Odd Beacon Interval Even Beacon Interval
Odd Beacon Interval Even Beacon Interval
B
╳
B
B B
╳M
M
M
M
52
Host A
Host B
odd beacon interval
Beacon window MTIM WindowActive window
odd beacon interval
even beacon interval
even beacon interval
Unicast Example
MTIM Data
ACKACK
53
Characteristics dominating awake
wake-up ratio < 1/2 sensibility
A PS host can receive a neighbor’s beacon once every two beacon intervals.
suitable for highly mobile environment
54
P2: Periodical-Fully-Awake-Interval
Basic Idea: In every T intervals, stay awake in one full
interval. wake-up ratio 1/T
compared to 1/2 of protocol 1
Two types of beacon intervals: Low-power interval Fully-awake interval (in every T intervals)
55
Example (T = 3)
Host A
Host B
T(=3) Beacon Intervals
Fully-awake
Fully-awake Low-power
Low-power
Low-power
Low-power
╳ ╳ ╳╳ ╳ ╳
T: Interval between the fully awake periods
A PS host can receive its neighbor’s beacon frame in every T = 3 beacon intervals
56
Definitions of Intervals Low-power interval:
active window + doze window AW = BW + MW
i.e., listening period = 0
Fully-awake interval: no doze window
i.e., AW = BI
very energy-consuming, so only appears once every T beacon intervals
57
P3: Quorum-Based Quorum Sets:
Two quorum sets always have nonempty intersection.
(used here to guarantee detectability) A matrix example:
n
nc1
c2
r1 r2
Host A’s quorum intervals
Host B’s quorum intervals
Non-quorum intervalsintersection
58
Example (2D matrix quorum)
0 1 2 3
4 5 6 7
8 9 10 11
12 13 14 15
0 1 2 3
4 5 6 7
8 9 10 11
12 13 14 15
Host A’s quorum intervals
Host B’s quorum intervals
Non-quorum intervals
Host A’ quorum intervals
1514131211109876543210
31302928272625242322212019181716
Group 1
Group 2
Host B’s quorum intervals
1514131211109876543210
31302928272625242322212019181716
Group 1
Group 2
Overlapping intervals
59
Overlapping Property Overlap no matter how clocks drift
demo ...
0 1 2 3
4 5 6 7
8 9 10 11
12 13 14 15
0 1 2 3
4 5 6 7
8 9 10 11
12 13 14 15
1514131211109876543210
1514131211109876543210
Host A’s quorum intervals
Host B’s quorum intervals
60
Quorum and Non-quorum Intervals
Quorum interval: AW = BI (i.e., fully awake)
Non-quorum interval: no beacon, only MTIM window AW < BI BW = 0, AW = MW
Beacon window MTIM WindowActive window
Beacon Interval Beacon IntervalQuorum Interval
Non-quorum Interval
61
Summary
Protocol Numbers of beacons per inter
val
Active ratio
Neighbor sensitivity
Dominating 1 1/2+BW/BI BI
Periodical 1 1/T T*BI/2
Quorum (2n-1)/n2 (2n-1)/n2 (n2/4) * BIBI: length of a beacon interval
AW: length of an active window
BW: length of a beacon window
MW: length of an MTIM window
T: interval between the fully awake periods
n: length of the square
62
Summary Identify the problems of PS mode in
IEEE 802.11 in multi-hop ad hoc networks. clock drifting, network-partitioning
Propose several PS protocols Connecting this problem to quorum
issue in distributed systems.
63
Sniff Scheduling for Power Saving in Bluetooth
64
Overview (cont.) Addressing
48-bit Bluetooth Device Address (BD_ADDR)
3-bit Active Member Address (AM_ADDR)
8-bit Parked Member Address (PM_ADDR)
Four operational modes: Active Sniff Hold Park
65
Bluetooth Networks Piconet
one master + at most 7 active slaves Scatternet
multiple piconets to form a larger network
Packets Exchange Scenario
MASTER
SLAVE 1
SLAVE 2
SLAVE 3
ACLSCO SCO SCO SCOACLACL ACL
67
Low-Power Sniff Mode A slave can enter the low-power sniff m
ode by setting a parameter (Tsniff, Nsniff_attempt, Dsniff)
in per slave basis
68
LMP_PDUs for Sniff
23 timing control flag Dsniff Tsniff Nsniff_attempt Nsniff_timeoutLMP_sniff_req
LMP_not_accepted
LMP_accepted
LMP_unsniff_req
4 op code reason
3 op code
24
LMP_sniff_req
LMP_accepted / LMP_not_accepted
initiating LM LM
LMP_sniff_req..
LMP_unsniff_req
LMP_accepted
.
.
69
Sniff Scheduling Problem How to determine the sniff parameters?
Goal: balancing power consumption and traffic need
Earlier Works naïvely adjust parameters in an
exponential way double/halve sniff interval or active window
whenever polling fails/succeeds
The placement of active windows of multiple slaves on the time axis is not addressed.
70
Design Goals consider multiple slaves together adaptively schedule sniff parameters
more accurate in determining the sniff-related parameters based on slaves’ traffic loads
include solutions of placing of active windows of sniffed slaves on the time axis
71
Proposed Architecture
LM Evaluator
LM Evaluator
LM Evaluator
Evaluator for Slave1
Evaluator for Slave2
Evaluator for SlavekResource
Pool
Scheduler
S1
SkSearching
S1
S2
Sk
Slave 1
Slave 2
Slave k
Master
.
.
.
.
.
.
Link Manager(LM)
S2
sniff related LMP packets
72
Tk,Nk,Dk: current sniff parameters for slave k. Uk: the slot utilization of slave k. Bk: the buffer backlog of slave k. Wk: a weighted value to indicate the current requirement of s
lave k.
Bmax is the maximum buffer space
Sk: the desired slot occupancy of slave k, which is the expected ratio of Nk / Tk.
0 < δ < 1 (to tolerate some unexpected traffic)
CalculatorX
Uk
Bk
CalculatorY
Sk Wk
the evaluator
73
Resource Pool (RP) Although time slots are an infinite seque
nce, we represent them as a sequence of 2-D matrices. each matrix M is of the size 2u × T time slots are viewed in a “row-major” way
The availability of M:
74
RP Example M’s size = 23 × 15 = 120
1 0 0 0 0
1
0
1
1
1
0
1
1 1 0 0
0 0 0 0
1 1 0 0
0 0 0 0
1 1 0 0
0 0 0 0
1 1 0 0
1 1 1 1 1
0
0
0
1
0
0
0
0 0 0 0
0 0 1 1
0 0 0 0
1 1 1 1
0 0 0 0
0 0 1 1
0 0 0 0
1 0 0 0 0
1
1
0
1
1
1
0
1 1 1 1
1 1 1 0
0 0 0 0
0 0 0 0
1 1 1 1
1 1 1 0
0 0 0 0
column
row
0
1
2
3
4
5
6
7
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
75
Example: to allocate a slot occupancy of 16/120(** Note: 16/120 = 8/60)
76
Example: to allocate a slot occupancy of 16/120(** Note: 16/120 = 4/30 = 2/15)
77
A Running Example 5 slaves Each slave initially has an equal occupa
ncy of 1/5 of the matrix M. We discuss two strategies:
longest sniff interval first shortest sniff interval first
78
Scheduling Policies:Longest Sniff Interval First(LSIF)
a) initial state (equal shares)
b) reduce S2 to 2/60
c) reduce S3 to 3/120
d) increase S4 to 6/30
S1 S1 S1 S2 S2S1S1S1S1S1S1S1
S1 S1 S2 S2S1 S1 S2 S2S1 S1 S2 S2S1 S1 S2 S2S1 S1 S2 S2S1 S1 S2 S2S1 S1 S2 S2
S2 S3 S3 S3 S4S2S2S2S2S2S2S2
S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4
S4 S4 S5 S5 S5S4S4S4S4S4S4S4
S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5
Round0 N1/T1 = N2/T2 = N3/T3 = N4/T4 = N5/T5 = 3/15
S1 S1 S1S1S1S1S1S1S1S1
S1 S1S1 S1S1 S1 S2 S2S1 S1S1 S1S1 S1S1 S1 S2 S2
S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4
S4 S4 S5 S5 S5S4S4S4S4S4S4S4
S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5
Round1 W2 = 0.18 < rlb => S2 = [(3/15)*0.18]/0.8 = 0.045 0.045 = 5/120 = 2/60 = 1/30 = 0/15 => (Ok' , Tk' , Nk' ) = (3, 60, 2)
S1 S1 S1S1S1S1S1S1S1S1
S1 S1S1 S1S1 S1 S2 S2S1 S1S1 S1S1 S1S1 S1 S2 S2
S4
S3
S4S4S4S4S4S4
S3 S3 S4
S4 S4 S5 S5 S5S4S4S4S4S4S4S4
S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5
Round2 W3 = 0.11 < rlb => S3 = [(3/15)*0.11]/0.8 = 0.028 0.028 = 3/120 = 1/60 = 0/30 = 0/15 => (Ok' , Tk' , Nk' ) = (5, 120, 3)
S1 S1 S1 S4 S4S1S1S1S1S1S1S1
S1 S1S1 S1 S4 S4S1 S1 S2 S2S1 S1 S4 S4S1 S1S1 S1 S4 S4S1 S1 S2 S2
S4 S4 S4 S4
S4
S4
S4S3
S4 S4 S4
S4 S4 S4
S4 S4 S4S3 S3
S5 S5 S5S5 S5 S5S5 S5 S5S5 S5 S5S5 S5 S5S5 S5 S5S5 S5 S5S5 S5 S5
Round3 W4 = 0.9 > rub => S4 = [(3/15)*0.9]/0.8 = 0.23 0.23 = 27/120 = 13/60 = 6/30 = 3/15 => (Ok' , Tk' , Nk' ) = (18, 30, 6)
79
S1 S1 S1 S2 S2S1S1S1S1S1S1S1
S1 S1 S2 S2S1 S1 S2 S2S1 S1 S2 S2S1 S1 S2 S2S1 S1 S2 S2S1 S1 S2 S2S1 S1 S2 S2
S2 S3 S3 S3 S4S2S2S2S2S2S2S2
S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4
S4 S4 S5 S5 S5S4S4S4S4S4S4S4
S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5
Round0 N1/T1 = N2/T2 = N3/T3 = N4/T4 = N5/T5 = 3/15
S1 S1 S1S1S1S1S1S1S1S1
S1 S1 S2S1 S1S1 S1 S2S1 S1S1 S1 S2S1 S1S1 S1 S2
S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4S3 S3 S3 S4
S4 S4 S5 S5 S5S4S4S4S4S4S4S4
S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5
Round1 W2 = 0.18 < rlb => S2 = [(3/15)*0.18]/0.8 = 0.045 0.045 = 5/120 = 2/60 = 1/30 = 0/15 => (Ok' , Tk' , Nk' ) = (3, 30, 1)
S1 S1 S1S1S1S1S1S1S1S1
S1 S1 S2S1 S1S1 S1 S2 S3S1 S1S1 S1 S2S1 S1S1 S1 S2 S3
S4S4S4S4S4S4S4S4
S4 S4 S5 S5 S5S4S4S4S4S4S4S4
S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5S4 S5 S5 S5
Round2 W3 = 0.11 < rlb => S3 = [(3/15)*0.11]/0.8 = 0.028 0.028 = 3/120 = 1/60 = 0/30 = 0/15 => (Ok' , Tk' , Nk' ) = (4, 60, 1)
S1 S1 S1S1S1S1S1S1S1S1
S1 S1 S2S1 S1S1 S1 S2 S3S1 S1S1 S1 S2S1 S1S1 S1 S2 S3
S4 S4 S4S4S4S4S4S4S4S4
S4 S4S4 S4S4 S4S4 S4S4 S4S4 S4S4 S4
S5 S5 S5S5 S5 S5S5 S5 S5S5 S5 S5S5 S5 S5S5 S5 S5S5 S5 S5S5 S5 S5
Round3 W4 = 0.9 > rub => S4 = [(3/15)*0.9]/0.8 = 0.23 0.23 = 27/120 = 13/60 = 6/30 = 3/15 => (Ok' , Tk' , Nk' ) = (5, 15, 3)
Scheduling Policies:Shortest Sniff Interval First(SSIF)
a) initial state (equal shares)
b) reduce S2 to 1/30
c) reduce S3 to 1/60
d) increase S4 to 3/15
80
Conclusions Proposed:
Power-saving protocols for IEEE 802.11-based multi-hop ad hoc networks
Sniff-scheduling schemes for Bluetooth-based piconets
References:1. T.-Y. Lin and Y.-C. Tseng, “An Adaptive Sniff Scheduling S
cheme for Power Saving in Bluetooth”, IEEE Personal Communications (to appear).
2. Y.-C. Tseng, C.-S. Hsu, and T.-Y. Hsieh, “Power-Saving Protocols for IEEE 802.11-Based Multi-Hop Ad Hoc Networks”, IEEE INFOCOM, 2002.