The Spectrum Sensing For Cognitive Radio Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

The frequency range used for Wireless Communications is 3 KHz to 300GHz which is known as radio spectrum. Due to the Rapid development in wireless usage FCC declared some frequencies as licensed spectrum and Remaining is left as unlicensed band. As more number of users are using the unlicensed band and some of licensed band frequencies, the traffic is heavily increased and we observe from below figure that there is limited usage in other licensed band frequencies. Therefore in order to provide service to all the users, FCC has been considering using the total available spectrum as efficiently as possible. In order to achieve this, Cognitive Radio (CR) Technology came into existence.

Cognitive Radio is an upcoming technology for the next generation Wireless Communications. The licensed band frequency users are primary users. When the primary user does not use the frequency band it can be used by secondary user. When primary user who has the right to use the frequency band comes into existence then secondary user should vacate the spectrum. Cognitive radio will perform spectrum sensing, dynamic spectrum sharing and handoff.

The Main challenge faced by the cognitive radio is that the unlicensed user should not interrupt the licensed users when they required using it. In order to accomplish this it should sense the signal faster.

Cognitive Radio functions:

The unused spectrum is called as the white space. The important function of the cognitive radio is to detect and select the unused frequency band and vacating the spectrum when the primary users want to use the spectrum.

Therefore the cognitive radio should perform these tasks:

Sensing and analyzing the spectrum

Managing and handing off the spectrum

Allocating and sharing the spectrum

The operating parameters (for example Modulation, Frequency, Transmission power etc) can be changed by Radio device according to the changes in environment. Cognitive Radio should know the information from the radio environment and adjust the operating mode to that environment.

Sensing and analyzing the spectrum means cognitive radio should continuously sense the spectrum; if it finds a free spectrum then CR can allocate it to secondary users. When primary user exists again in the spectrum then secondary user should vacate in order to avoid the interference.

Managing and handing off the spectrum means CR should be able to select the best frequency from the available frequencies and allocate that particular one to secondary user. CR should also able to shift the secondary user between the different frequency bands in order to provide good service when primary user comes. This can be determined by the channel capacity which include many other factors like holding time, channel error rate, path loss, noise and interference levels etc. The CR functions are only equipped in secondary base stations and users but not generally in primary base stations and users.

Due to natural disaster communications network can be damaged, so for emergency also CR network can be used. CR network can also provide public safety and home land security as it can easily adapt to other frequencies.


In dynamic spectrum management the first critical step is Spectrum Sensing and Analyzing.

Spectrum sensing consists of three different aspects:

Figure: Different techniques for spectrum sensing

Non cooperative system is for detection of primary transmitter. In Cooperative system, CR cooperates with other CR’S whenever there is a need of high sensitivity of users and also used to reduce effects of noise uncertainties, shadowing, multipath fading, false alarms. In Interference based sensing, users of CR will work in approach of spectrum underlay.

Detection of Primary Transmitter: It is used to detect extra spectrum resources that are available and to have comparison with several detection techniques.

Energy Detector

Cyclostationary Feature Detector

Coherent Detection using Matched Filtering

Energy Detector: It is the most commonly used method since it does not require any primary signal knowledge.

Consider the recieved signal that is minimized by the hypothesis model

H0: z(k)=n(k)

H1: z(k)=h*x(k)+n(k)

x(k) is detected at the secondary user local receiver that tells about primary user.

n(k) : AWGN.

h : Channel Gain at the secondary user receiver from the primary user transmitter

H0: Null hypothesis means absense of primary user.

H1: Primary user presence.

The energy detector detection statistics is given as average energy of N observed samples

The presence of primary user in the spectrum is determined by evaluating T (detection statistics) and comparing with (pre-determined threshold). The detector performance is characterised as detection probability PD and false alarm probability PF. The probability that the test correctly decides H1 denoted by PD.

The result is H0 but if the hypothesis test decides it to be H1 then it is said to be false alarm probability PF.

The high probability detection PD and low probability false alarm PF are the characteristics of good detector.

But in this detection, primary source cannot be differentiated from other signals. Also there is a chance of high probability of false alarm that is caused by noise uncertainity and low SNR regimes unreliability.

Cyclo Stationary Feature Detection:

One more method in non cooperative system for the primary signal detection is Cyclostationary Feature Detection. In this detection the modulated signals are coupled with cyclic prefixes, hopping sequences, repeated spreading, pulse trains, sine wave carriers. Since the mean and auto correlation of modulated signals exhibits periodicity this method is called as cyclostationary. The main advantage of this method is it differentiates between noise and primary signals and also differentiates among different types of transmissions and primary signals. In energy detector the hypothesis test is done in time domain but the cyclostationary detector performs transformation from time to frequency domain and then performs hypothesis test.

Cyclic auto correlation function of a received signal y(t) is

* denotes complex conjugation

E[.] expectation operation

cyclic frequency

Cyclic spectrum density function is given as


Coherent detection using matched filtering:

The other classification of non cooperative system is is a linear filter that maximizes SNR which requires the primary signal knowledge such as packet format, pulse shape and modulation scheme. The information of primary signal is generally stored in CR memory and if it is not accurate, results could be poor. The main advantage is it requires less time to achieve high processing gain because of coherency.

There are two hypotheses in coherent detection

Xp(k) : pilot tone

is fraction of energy allocation for pilot tone

x(k) are the signals which are desired and assumed to be Orthogonal to Xp(k) (pilot tone)

n(k) is AWGN

As N increases, T value under Hypothesis H1 is greater than under H0. The presence of primary user can be decided by predetermined dectection threshold which is compared to T value.