1

Spectrum Sensing

Marjan Hadian

2

Outline

• Cognitive Cycle• Enrgy Detection• Matched filter• cyclostationary feature detector• Interference Temperature• Spectral Estimation• Hidden node problem• Cooperative detection• detection methods

– log-likelihood combining– weighted gain combining

3

Cognitive Cycle

Mitola calls cognitive radio cycle: cognitive radio continually observes the environment, orients itself, creates plans, decides, and then acts

4

Spectrum Sensing:A cognitive radio monitors the available spectral

bands,captures their information, and detects the spectrum holes.

• frequencies usage.• mode identification.

5

6

• Enrgy Detection

Where T calculated from:

most important problem of this, is which one called SNR wall. This problem comes from uncertainty.

SNR wall is a minimum SNR below which signal cannot be detected and formulas no longer holds

7

• Matched filter

it maximizes SNR. For implementation of matched filter cognitive radio has a priori knowledge of modulation type, pulse shaping.

• cyclostationary feature detectorThe main advantage of the spectral correlation

function is that it differentiates the noise energy from modulated signal energy.

8

Interference TemperatureAs additional interfering signals appear the noise floor

increases and then unlicensed devices could use that particular band as long as their energy is under mention noise floor

where Joules per Kelvin

9

Spectral Estimation

• parametric spectral estimation• Non-parametric spectral estimation

Periodogram Spectral Estimator (PSE) Blackman-Tukey Spectral Estimator (BTSE) Minimum Variance Spectral Estimator (MVSE) Multi taper Method (MTM) Filter Bank Spectral Estimator (FBSE)

10

Hidden node problem

Traditional detection problem: (a) Receiver uncertainty and (b) shadowing uncertainty[5]

11

Cooperative detection

• prevent the hidden terminal problem also mitigate the multipath fading and shadowing effect

• Information from multiple SUs are incorporated for primary user detection.

• Implementation Centralized manner distributed manner

12

How SU provide its observation to other nodes?!

• This transmission can overlap to the air interfaces already present in the environment, so it can change the nature of observations and make new problems. In order to solve this problem several solutions suggested :

two distinct networks are deployed separately the sensor network for cooperative spectrum sensing and

the operational network for data transmission. This method implemented in central manner[5]

Sharing the analysis model in an off-line method when in the environment no SUs is observing the radio scene[1]

13

Without consideration of exchanging method, we assume that the observation of SU i is due to its position and to the state of radio source, but not to the observation of other SU j and .Thus we assume that, independent measurements for each SUs is presented either in a centralized or distributed manner. Now we review two detection methods:

• log-likelihood combining• weighted gain combining

14

• log-likelihood combiningAssume that is the vector of SUs energy detector

output, then we can write likelihood ratio test(LRT) as:

• weighted gain combining:

where and

15

Thanks for your attention.Questions?