cognitive radio communication wireless systems

cognitive radio communication

1. COGNITIVE RADIO
DEFINITION:
A radio that can change its transmitter parameters
based on interaction with the environment in which it
operates.

2. SPECTRUM UTILIZATION

3. BASIC TERMS IN COGNITIVE RADIO
Spectrum gap, spectrum hole:
Spatially & temporally unused part of the radio
spectrum, which is considered for use by CR
Primary user (PU):
Licensed user/privileged user of a frequency band
 Secondary user (SU) :
Opportunistic user of a frequency band

4. SPECTRUM HOLE

6. SPECTRUM SENSING
Analyzing a radio spectrum environment to identify
temporarily vacant spectrum and use it.
Why SPECTRUM SENSING??
To avoid interference with PU
 Guarantee a good quality of service

7. SPECTRUM SENSING TECHNIQUES

8. TRANSMITTER DETECTION
Whether the signal from a primary transmitter is locally
present in a certain spectrum or not

Different approaches of Transmitter Detection :
 Energy detection
Matched filter detection
Eigen Value Based detection
Cyclostationary detection

9. ENERGY DETECTION
Energy detection method does not require any prior
information of the signal.
R(t)

10. Measures the energy of available radio resource and
compare it against a predefined threshold level.
 If the measured energy falls below the defined threshold
level spectrum is marked as available.
When the measure energy level is above the defined
threshold, it’s considered as occupied.

11. ADVANTAGES
It does not require any prior information of the signal
optimum detection technique if the primary user signal is
not known
 low implementation and computational complexities
DISADVANTAGES
 Cannot distinguish between Signal and Interference
 No reliable detection beyond SNR wall

12. MATCHED FILTER DETECTION
It is a method of detection when the transmitted signal is
well-known
Detection of PU based on prior knowledge (bandwidth,
operating frequency , modulation type etc).
r(t)
Decide H0 or H1

13. ADVANTAGES:
The detector does not suffer from the “SNR wall”
problem
It takes moderate time for execution
DISADVANTAGES:
 It necessitates the faultless knowledge of primary users
signalling features
Complexity is large
Huge power consumption

14. EIGEN VALUE BASED DETECTION
Known as Blind Spectrum Sensing—a priori information
about the signal is not needed

15. 2 TYPES OF EIGEN VALUE DETECTION
1.MAXIMUM-MINIMUM EIGENVALUE DETECTION(BLIND)
Compares the ratio of the maximum eigenvalue
and the minimum eigenvalue with a threshold
2. ENERGY WITH MINIMUM EIGENVALUE
DETECTION(SEMI-BLIND)
Compares the ratio of the average energy and
the minimum eigenvalue with a threshold

16. ADVANTAGES
Prior knowledge about the signal is not needed
Implementation is simple
Execution time is moderate
DISADVANTAGES
Ability to distinguish between signal and interference is
less.

17. COMPARISON
Executio
n time
Ability to
distinguish
Signal&
Interference
Prior
Information
Implementat
ion
Matched
Filter
Moderate
time
Medium Necessary Complex
Energy
Detection
Less timeLow Not
Necessary
Simple
Features
detection
More
time
High Necessary Very
Complex
Eigen
Values
detection
Moderate
time
Low Not
Necessary
Simple

18. ENERGY DETECTION

19. FLOW CHART
START
IInitialise No. of samples N and fix Pfa value
Set Threshold using Neyman Pearson rule
T=gammaincinv(1-Pfa(i),N)m
Generate Primary user signal x
Generate Noise n
A

20. A
Secondary user received signal
Y=x+n
Compute energy of the signal
E=sum(abs(Y.^2))
If E>T PU detected
No Primary User
YES
NO

21. MATCHED FILTER BASED DETECTION

22. FLOW CHART
START
IInitialize SNR, signal power, noise power
Set Threshold using Neyman Pearson rule
T=sqrt(npower*mfgain*snrthreshold)
Generate Primary user signal
s(t) and noise n(t)
Secondary user received
signal
x=s(t)+n(t)
A

23. A
Matched filter Output
Y=mf ‘ *x
Take Z=real(Y)
If Z>T PU detected
No Primary User
YES
NO

24. MAXIMUM-MINIMUM EIGENVALUE (MME)
DETECTION
ALGORITHM:
Step 1: Find the Sample Covariance Matrix of the
received signal.
Where Ns is the number of collected samples.

25. Step 2: Compute the Maximum and Minimum
Eigenvalues of the Matrix Rx(Ns) , λmax and λmin
Step 3:Decision:
If λmax / λmin > ᵞ1 –signal exists
else—signal doesnot exist
Where ᵞ1 is the threshold

26. THRESHOLD
The threshold depends on the probability of false alarm
(Pfa), number of samples ( Ns), number of receiver (M) and
the smoothing factor (L).

27. THANK YOU