Presentation

College of Engineering

Resource Management in

Wireless Networks

Resource Management in

Wireless Networks

Anurag Arepally

Major Adviser : Dr. Robert Akl

Department of Computer Science and Engineering

Anurag Arepally

Major Adviser : Dr. Robert Akl

Department of Computer Science and Engineering

04/12/23 2/56

OutlineOutline

• Wireless Networks History

• Resource Management Issues

• Call Admission Control in 3G UMTS WCDMA Systems

• Dynamic Channel Assignment in IEEE 802.11 Systems

• Future Work

• Wireless Networks History

• Resource Management Issues

• Call Admission Control in 3G UMTS WCDMA Systems

• Dynamic Channel Assignment in IEEE 802.11 Systems

• Future Work

04/12/23 3/56

Wireless Networks History(1/8)


• Mobile Communications

• Third Generation Partnership Project (3GPP)

• UMTS / WCDMA Overview

• IEEE 802.11

• WLAN Overview


• Third Generation Partnership Project (3GPP)

• UMTS / WCDMA Overview

• IEEE 802.11

• WLAN Overview

04/12/23 4/56




• Access Techniques

• Early 90’s saw the introduction of two access techniques

• TDMA Interim Standard 54 and IS-136 (updated version)

• CDMA IS-95 (code division multiple access)

• 3GPP introduces WCDMA (wideband code division multiple access) based on CDMA


• Access Techniques

• Early 90’s saw the introduction of two access techniques

• TDMA Interim Standard 54 and IS-136 (updated version)

• CDMA IS-95 (code division multiple access)

• 3GPP introduces WCDMA (wideband code division multiple access) based on CDMA

04/12/23 5/56



• 3GPP formed in late 90’s

• 3GPP develops standards for 3G networks based on global system for mobile communications (GSM)

• 3GPP2 develops standards for 3G networks based on CDMA IS-95

• 3GPP formed in late 90’s

• 3GPP develops standards for 3G networks based on global system for mobile communications (GSM)

• 3GPP2 develops standards for 3G networks based on CDMA IS-95

04/12/23 6/56



• 3GPP Releases

• Release ’99

• Voice and video use circuit switched network

• SMS, WAP, and MMS use packet switched network

• Release 5 introduced all IP-network

• Release 6 – mobile TV

• Release 7 and long term evolution

• 3GPP Releases

• Release ’99

• Voice and video use circuit switched network

• SMS, WAP, and MMS use packet switched network

• Release 5 introduced all IP-network

• Release 6 – mobile TV

• Release 7 and long term evolution

04/12/23 7/56



• Universal mobile telecommunications system (UMTS)

• Proposed by ETSI

• Backward compatible with 2G networks

• UMTS Terrestrial Radio Access (UTRA)

• Universal mobile telecommunications system (UMTS)

• Proposed by ETSI

• Backward compatible with 2G networks

• UMTS Terrestrial Radio Access (UTRA)

04/12/23 8/56



• WCDMA is the preferred access technique for 3G UMTS networks

• Main features of WCDMA

• Based on direct sequence CDMA

• Frequency spectrum of 5 MHz

• Multiplexing is done both in frequency (FDD) and time (TDD)

• WCDMA is the preferred access technique for 3G UMTS networks

• Main features of WCDMA

• Based on direct sequence CDMA

• Frequency spectrum of 5 MHz

• Multiplexing is done both in frequency (FDD) and time (TDD)

04/12/23 9/56



• IEEE 802.11 committee formed in 1990 for wireless LANs (WLAN)

• Unlicensed industrial, scientific, and medical bands – 915 MHz, 2.4 GHz, and 5 GHz

• 802.11a (1999) - 5 GHz, 54 Mbps

• 802.11b (1999) - 2.4 GHz, 11 Mbps

• 802.11g (2003) - 2.4 GHz, 54 Mbps

• IEEE 802.11 committee formed in 1990 for wireless LANs (WLAN)

• Unlicensed industrial, scientific, and medical bands – 915 MHz, 2.4 GHz, and 5 GHz

• 802.11a (1999) - 5 GHz, 54 Mbps

• 802.11b (1999) - 2.4 GHz, 11 Mbps

• 802.11g (2003) - 2.4 GHz, 54 Mbps

04/12/23 10/56



• WLAN

• Data Transmission

• DSSS (Direct Sequence Spread Spectrum)

• FHSS (Frequency Sequence Spread Spectrum)

• 802.11 MAC uses CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance)

• WLAN

• Data Transmission

• DSSS (Direct Sequence Spread Spectrum)

• FHSS (Frequency Sequence Spread Spectrum)

• 802.11 MAC uses CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance)

04/12/23 11/56

Resource Management IssuesResource Management Issues

• Capacity of the cellular and wireless networks

• Quality of Service (QoS)

• Grade of Service (GoS)

• Different models and approaches have been proposed

• Demand for wireless internet access

• Capacity of the cellular and wireless networks

• Quality of Service (QoS)

• Grade of Service (GoS)

• Different models and approaches have been proposed

• Demand for wireless internet access

04/12/23 12/56

Call Admission Control(CAC)

Call Admission Control(CAC)

• 3G UMTS WCDMA networks

• Voice, video, pictures form different classes of services

• Global CAC

• Optimized local CAC

• Modeling and Simulations

• Conclusions

• 3G UMTS WCDMA networks

• Voice, video, pictures form different classes of services

• Global CAC

• Optimized local CAC

• Modeling and Simulations

• Conclusions

04/12/23 13/56

Global CAC(1/6)

Global CAC(1/6)

• Multi-cell UMTS networks

• Feasible call configuration

• Call arrival and admission module

• Average interference

• Actual interference

• Call removal module

• Multi-cell UMTS networks



• Average interference

• Actual interference


04/12/23 14/56

Global CAC(2/6)

Global CAC(2/6)


for i = 1, …, M, g = 1, …, G,

where W bandwidth

Rg information rate in bits / s

Sg received signal

Vg activity factor

No noise spectral density


for i = 1, …, M, g = 1, …, G,

where W bandwidth

Rg information rate in bits / s

Sg received signal

Vg activity factor

No noise spectral density

, , ,1 1, 1

,

g

gb

G M Go g

o i g g j g g ji g gg j j i g

S

RE

I SN n v n v v

Wκ

= = ≠ =

⎛ ⎞=⎜ ⎟

⎡ ⎤⎝ ⎠ + + −⎢ ⎥⎣ ⎦∑ ∑ ∑

04/12/23 15/56

Global CAC (3/6)

Global CAC (3/6)

( )

0

1 ggeff

g g g

RWc

R S Nτ⎡ ⎤

= −⎢ ⎥⎢ ⎥⎣ ⎦

is the minimum signal-to-noise ratio

gτ

is the maximum signal power

Sg

the number of users in BS i for given service g

,ni g

( ), , ,

1 1, 1

G M Gg

i g g j g g ji g g effg j j i g

n n cν ν κ ν= = ≠ =

+ − ≤∑ ∑ ∑

Feasible call configuration is a set of calls n satisfyingFeasible call configuration is a set of calls n satisfyingthe above equations, for all services g = 1,…,Gthe above equations, for all services g = 1,…,G

This is for perfect power control (PPC).This is for perfect power control (PPC).

wherewhere

04/12/23 16/56

Global CAC (4/6)

Global CAC (4/6)


for i = 1,…,M, g = 1,…,G.


for i = 1,…,M, g = 1,…,G.

( ) ( ) ( ) ( ), , , ,g

i g i g i g effC t n t I t c= + ≤

( ) ( ) ( ), , , ,i g i g i gt t v tρ λ= +

04/12/23 17/56

Global CAC (5/6)

Global CAC (5/6)

• Average Interference

for i = 1,…,M, g = 1,…,G.

• Average Interference

for i = 1,…,M, g = 1,…,G.

, , ,1

,M

i g j g ji gj

I n F=

=∑

( ) ( ) ( ) [ ], , ,1

, , ,M

i g i g j gj

C t n t n t F j i g=

= + ∑

04/12/23 18/56

Global CAC (6/6)

Global CAC (6/6)

• Actual Interference

for i = 1,…,M, g = 1,…,G .


• Actual Interference

for i = 1,…,M, g = 1,…,G .


[ ]( ) [ ]( ) ( ),, , , , 1 ,ji g kU j i g t U j i g t U= − +

( ) ( ) [ ]( ), ,1,

, , ,M

i g i gj j i

C t n t U j i g t= ≠

= + ∑

04/12/23 19/56

Optimized Local CAC(1/6)

Optimized Local CAC(1/6)

• Admissible call configuration

• Calculation of N

• Theoretical Throughput

• Simulator Model



• Simulation Results



• Theoretical Throughput

• Simulator Model



• Simulation Results

04/12/23 20/56

Optimized Local CAC (2/6)



for i = 1,…,M and g = 1,…,G,

where denotes maximum # of calls

with service g in cell i.

• Blocking probability for cell i with service g is


for i = 1,…,M and g = 1,…,G,

where denotes maximum # of calls

with service g in cell i.

• Blocking probability for cell i with service g is

, ,i g i gn N≤,i gN

( )

,

,

,

,, , ,

,0

!, ,

!

i g

i g

N

i g

i gi g i g i g N

ki g

k

T

NB B T N

T

k=

= =

∑

04/12/23 21/56



where is the Erlang traffic

in cell i with service g.


where

where is the Erlang traffic

in cell i with service g.


where

( ),

,1i g

i

ii g

Tq

ρ

μ=

−

( ) ( ) ( ){ }, , , , ,1 1

, , 1 ,M G

i g i g i g i g i gi g

H B Bρ λ λ ρ λ= =

= − − −∑∑B

: vector of blocking probabilities

: matrix of call arrival ratesλB

04/12/23 22/56



maxN

subject to , ,( , ) ,i g i g gB T N β≤

( ), , ,

1 1, 1

,G M G

gi g g j g g ji g g eff

g j j i g

N v N v v cκ= = ≠ =

+ − ≤∑ ∑ ∑

for i = 1, …, M .

( , ),H , ρ λB

The above optimization problem is solved

offline to obtain the values of N.

The above optimization problem is solved

offline to obtain the values of N.

04/12/23 23/56



maxλ , N

subject to, ,( , ) ,i g i g gB T N β≤

( ), , ,

1 1, 1

,G M G

gi g g j g g ji g g eff

g j j i g

N v N v v cκ= = ≠ =

+ − ≤∑ ∑ ∑

for i = 1, …, M .

( , ),H , ρ λB

• Theoretical Throughput• Theoretical Throughput

04/12/23 24/56



• Simulator model



• Simulator model



04/12/23 25/56

SimulationSimulation

• Network configuration

• COST-231 propagation model

• Carrier frequency = 1800 MHz

• Average base station height = 30 meters

• Average mobile height = 1.5 meters

• Path loss coefficient, m = 4

• Shadow fading standard deviation, σs = 6 dB

• Bit energy to interference ratio threshold, τ = 7.5 dB

• Interference to background noise ratio, I0/N0 = 10 dB

• Activity factor, v = 0.375

• Network configuration

• COST-231 propagation model

• Carrier frequency = 1800 MHz

• Average base station height = 30 meters

• Average mobile height = 1.5 meters

• Path loss coefficient, m = 4

• Shadow fading standard deviation, σs = 6 dB




04/12/23 26/56

Simulation ResultsSimulation Results

• Processing gain, W/Rg

• 24.08 dB for spreading factor = 256







• Processing gain, W/Rg








04/12/23 27/56

Three Mobility ModelsThree Mobility Models

,ii gq

probability that a call with service g in progress in cell i departs from the network.

,ij gq

,i gq

probability that a call with service g in progress in cell i remains in cell i after completing its dwell time.

probability that a call with service g in progress in cell i after completing its dwell time goes to cell j. It’s equaled zero (=0) if cell i and j are not adjacent.

No MobilityNo Mobility

Low MobilityLow Mobility

High MobilityHigh Mobility

04/12/23 28/56

Simulated Network CapacitySimulated Network Capacity

04/12/23 29/56

UMTS Throughput Optimization with SF = 256


04/12/23 30/56

Average Throughput in each cell for SF = 256


04/12/23 31/56



04/12/23 32/56



04/12/23 33/56



04/12/23 34/56



04/12/23 35/56



04/12/23 36/56



04/12/23 37/56

Conclusions of CACConclusions of CAC

• Different spreading factors and various mobility scenarios

• Computational complexity for global CAC using average and actual interference is O(MG) and O(M2G)

• Optimized local CAC is O(1)

• Performance difference is less than 5%

• Different spreading factors and various mobility scenarios

• Computational complexity for global CAC using average and actual interference is O(MG) and O(M2G)

• Optimized local CAC is O(1)

• Performance difference is less than 5%

04/12/23 38/56

Dynamic Channel Assignment in IEEE 802.11 systems

Dynamic Channel Assignment in IEEE 802.11 systems

• Channel interference

• Overlapping Channel interference factor

• Dynamic channel assignment

• Analysis of simulation results

• Conclusions

• Channel interference

• Overlapping Channel interference factor

• Dynamic channel assignment

• Analysis of simulation results

• Conclusions

04/12/23 39/56

Channel InterferenceChannel Interference

• Two Types

• Adjacent channel interference

• Co-channel interference

• Overlapping channel interference factor

• Two Types

• Adjacent channel interference

• Co-channel interference

• Overlapping channel interference factor

04/12/23 40/56

Channel InterferenceChannel Interference

04/12/23 41/56

Dynamic Channel Assignment(1/3)


be the set of neighboring APs to AP i

is the overlapping channel factor

is the distance between AP i and AP j

is the channel assigned to AP i

is the interference that AP j causes on AP i

is the total number of available channels

is a function that captures the attenuation loss

is a pathloss exponent

is the transmit power of AP i

is the cardinality of A_i

is the overlapping channel interference factor

between AP i and AP j

be the set of neighboring APs to AP i

is the overlapping channel factor

is the distance between AP i and AP j

is the channel assigned to AP i

is the interference that AP j causes on AP i

is the total number of available channels

is a function that captures the attenuation loss

is a pathloss exponent

is the transmit power of AP i

is the cardinality of A_i

is the overlapping channel interference factor

between AP i and AP j

iΑ

ijd

iF

ijI

iP

iQ

ijw

c

kLoss

m

04/12/23 42/56



• Dynamic channel assignment problem is given as :

min

subject to

if

otherwise,

for

• Dynamic channel assignment problem is given as :

min

subject to

if

otherwise,

for

1 | |ij i jw F F c= − −

1

,iQ

ijj

I=∑

iF

( ),,

ij jij

ij

w PI

Loss d m=

,ij ∈Α { }1,..., .iF K∈

0,ijw ≥0=

04/12/23 43/56



• We use two versions for analysis of our algorithm

• Algorithm I (pick rand)

• Algorithm II (pick first)

• We use two versions for analysis of our algorithm

• Algorithm I (pick rand)

• Algorithm II (pick first)

04/12/23 44/56

Analysis of Simulation Results(1/10)


• Signal level maps• Signal level maps

04/12/23 45/56



• Signal level maps• Signal level maps

04/12/23 46/56

Analysis of Simulation Results (3/10)


• Dynamic Channel Assignment for WLAN with 4 APs


04/12/23 47/56



• Channel Assignment map for WLAN with 4 APs


04/12/23 48/56





04/12/23 49/56





04/12/23 50/56





04/12/23 51/56





04/12/23 52/56





04/12/23 53/56





04/12/23 54/56

ConclusionsConclusions

• WLAN consisting of 4, 9, 16, and 25 AP s

• Default factory settings ( all AP’s are assigned same channel number)

• Algorithm I (pick rand) and II (pick first)

• Our results show an improvement by a factor of 4 or 6 dBm

• WLAN consisting of 4, 9, 16, and 25 AP s

• Default factory settings ( all AP’s are assigned same channel number)

• Algorithm I (pick rand) and II (pick first)

• Our results show an improvement by a factor of 4 or 6 dBm

04/12/23 55/56

Future workFuture work

• 4G the next revolutionary technology and WLAN complementing WCDMA will lead to integrated wireless networks.

• Dynamic load balancing

• 4G the next revolutionary technology and WLAN complementing WCDMA will lead to integrated wireless networks.

• Dynamic load balancing

04/12/23 56/56

Thank You!!Thank You!!

Questions?Questions?

Presentation

Documents

g umts wcdma networks

db interference

dwell time g

admission module fori

wireless internet access

configuration fori

db activity factor

admission control cac