radio resource management in ofdma-crn considering primary … · 2011-12-08 · kanagaraj n. n.,...

Cape Town, South Africa 12–14 December 2011

Dhananjay Kumar, M.E., M.Tech., Ph.D. Anna University, Chennai dhananjay@annauniv.edu

Kanagaraj N. N., B.Tech.

Alcatel-Lucent India Limited, Chennai

kanagaraj.nn@alcatel-lucent.com

Radio Resource Management in OFDMA-CRN Considering Primary

User Activity and Detection Scenario

ITU Kaleidoscope 2011 The fully networked human?

Innovations for future networks and services

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Outline

Cognitive Radio Environment

Real-time Observations & Requirement Analysis

Learning in Cognitive Radio Networks

Resource Allocation in Multi-carrier Systems

Problem Formulation in OFDMA-CRN

Simulation Results

Conclusion & Future Works

Cognitive Radio Environment

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

BASE STATION

CR USER 1

Cognitive Network (Secondary)

CR USER 3

CR USER 1

CR USER 4

CR USER 2

MESH ROUTER 1 MESH ROUTER 3

MESH ROUTER 2

PRIMARY USER 2

PRIMARY USER 3

PRIMARY USER 1

TV Broadcast Network

(Primary)

PRIMARY USER 4

PRIMARY USER 5

IEEE 802.11n Access Point (ISM Band)

Wireless LAN (Primary)

Wireless Mess Network (Primary)

Cognitive Link

Primary Link

Opportunity in 900 MHz Band

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Measurements taken at Adyar, Chennai on 16th August, 2011

Fig.2. At 12:24:44 p.m.

Fig.1. At 12:14:12 p.m.

Fig.3. At 12:49:13 p.m.

Spectrum Usage in 2.4 GHz Band

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Fig.1. At 11:30:43 a.m.

Fig.2. At 11:39:31 a.m.

Fig.3. At 11:39:51 a.m.

Fig.4. At 11:40:39 a.m.

On 24th August, 2011 (Adyar, Chennai)

Learning in Cognitive Radio Network

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Learning Algorithms

PARAMETERS RULES ENGINE

KNOWLEDGE

WAVEFORM INTERNAL STATE

EXTERNAL STATE

CONTROL

OPERATIONAL FUNCTIONS

SENSORS

RADIO SYSTEM

TX TX RX

WIRELESS ENVIRONMENT

RADIO COMMUNICATION CHANNELS

Learning in a Broader Perspectives

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Ref: Joseph Gaeddert et al., “Applying Artificial Intelligence to the Development of a Cognitive Radio Engine”, http://wireless.vt.edu/archives/download/ApplyingArtificialIntelligence.pdf

DOMAIN KNOWLEDGE BASE

CASE MEMORY

RADIO ENVIRONMENT

MAP(REM) OBSERVATION

HMM, DATA MINING

GENETIC ALGORITHMS

SEARCH BASED

LEARNING

NEURAL NETWORK

COOPERATIVE LESRNING

Prediction, planning

Parameter optimization,

tradeoff

Collaboration with other nodes

Learning and Reasoning

Situation Awareness and

Information Base

Decision and Adaptation

KNOWLEDGE BASED

LEARNING

CASE BASED

LEARNING

Resource Allocation in Multi-carrier Base Cognitive Radio Networks

Is situation-aware learning is enough in multi-carrier systems?

Resource allocation

Sub-carrier selection

Power allocation

Resource optimization

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Resource Allocation in OFDMA-CRN: Problem Formulation

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Objective Function:

Optimization Methods

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Optimization Methods cont..

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Sub-carrier and Power Allocation

(SPA) Algorithm

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Start

Categorize the secondary users into different groups

Initialize (λ1, λ2, . . .λN)

k=1(subcarrier)

Compute P*g, k for all groups

Pick the group g* that gives the minimum of Dk(λ)

Assign the sub-carrier to the group g*

Allocate power for the group g*

Set Pg*, k = P*g*, k and Pg, k =0 for all g ≠ g*

Increment k

Stop

Is k<K Yes

No

Main Simulation Parameters

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Parameters Description Value

N Number of primary users 2

K Number of OFDM sub-carriers 128

B Maximum spectrum hole 10MHz

G Number of groups 10

|Mg| Number of secondary users in

each group g

4

Effects of Varying Loss Parameters

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10.4

0.41

0.42

0.43

0.44

0.45

0.46

0.47

0.48

0.49

Phi (with fixed beta = 0.03)

Data

rate

(bps p

er

Hert

z)

C1 = 0.0

C1 = 0.2

C1 = 0.4

C1 = 0.6

C1 = 0.8

C1 = 1.0

Effects of Varying Detection Parameters

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10.44

0.46

0.48

0.5

0.52

0.54

0.56

0.58

0.6

Beta (with fixed phi = 0.04)

Data

rate

(bps p

er

Hert

z)

C2 = 0.0

C2 = 0.2

C2 = 0.4

C2 = 0.6

C2 = 0.8

C2 = 1.0

Sum Data Rates in Two Proposed Scenarios

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

2 3 4 5 6 7 8 9 100.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

No. of groups (secondary users)

Data

rate

(bps p

er

Hert

z)

Scenario 1(phi=0.08,beta=0.05)

Scenario 2(phi=0.08,beta=0.05)

Scenario 1(phi=0.04,beta=0.03)

Scenario 1(phi=0.04,beta=0.03)

Throughput Comparison

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

2 3 4 5 6 7 8 9 1016

17

18

19

20

21

22

23

24

No. of groups (secondary users)

Thro

ughput(

Mbps)

Scenario 1 (phi=0.08,beta=0.05)

Scenario 2 (phi=0.08,beta=0.05)

Scenario 1 (phi=0.04,beta=0.03)

Scenario 2 (phi=0.04,beta=0.03)

Assessment of Processing Delay

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

2 3 4 5 6 7 8 9 100.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

0.022

No. of groups (secondary users)

Pro

cessin

g d

ela

y(s

econds)

Scenario 1

Scenario 2

Conclusion & Future Work

Analysed the spectrum availability based on real-time measurements.

Discussed the learning scenarios

Defined the objective function and optimized it analytically.

Developed Sub-carrier and Power Allocation (SPA) Algorithm for OFDMA-CRN.

Analysis & simulation of the effect of both issues i.e. primary user activity and detection

Unification of learning algorithm with sub-carrier allocation in OFDMA-CRN is our future work.

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services

Thank You

Cape Town, South Africa, 12-14 December 2011 ITU Kaleidoscope 2011 – The fully networked human? Innovations for future networks and services