Spectrum Sensing :

Through spectrum sensing, CR can obtain necessary information about surrounding radio environment. These include the presence of primary users and appearance of spectrum holes. Only with this information CR can adapt its transmitting and receiving parameters, like transmission power, frequency, and modulation schemes in order to achieve efficient spectrum utilization. Therefore, spectrum sensing and analysis is the first critical step towards dynamic spectrum management. In this section, we will discuss two important aspects of spectrum sensing, Interference temperature and Spectrum hole detection.

Interference Temperature: In order to utilize the spectrum efficiently, it is necessary to detect the presence of primary users and to decide the available frequency band. Traditionally such a decision is based on measurement of transmitted power of device and this power should not be more than a prescribed noise floor at a certain distance from the transmitter. However, due to the increased mobility and appearance of radio frequency (RF) emitters (working as interference) puts a limit on the accurate measurement of transmitted power. To address this issue, FCC Spectrum Policy Task Force has proposed a new metric on interference assessment, the interference temperature, to enforce an interference limit perceived by receivers. The interference temperature is a measure of the RF power available at a receiving antenna to be delivered to a receiver, reflecting the power generated by other emitters and noise sources. More specifically, it is defined as the temperature equivalent to the RF power available at a receiving antenna per unit bandwidth.

$T_I (f_c, B) = \frac{P_I(f_c, B)}{kB}$

Where, $P_I$ is the average interference power in Watts centered at $f_c$, covering bandwidth B measured in Hz, and K Boltzmann’s constant is 1.38 Joules per degree Kelvin. With the concept of interference temperature, FCC further established an interference temperature limit, which provides a maximum amount of tolerable interference for a given frequency band at a particular location. Any unlicensed secondary user (transmitter) using this band must guarantee that their transmission plus the existing noise and interference must not exceed the interference temperature limit at a licensed user (receiver). If a regulatory body sets an interference temperature limit T for a particular frequency band with bandwidth B, then the secondary transmitters has to keep the average interference below $KBT$. Therefore, the interference temperature serves as a limit on potential RF energy that could appear on that band and is used to protect licensed primary users from harmful interference due to unlicensed secondary users.

Spectrum Hole Detection : Spectrum sensing enables the capability of a CR to determine the spectrum availability and interference status. When a certain frequency band is detected as not being used by the primary licensed user at a particular time at a particular position, secondary users can utilize this spectrum, i.e. there exists spectrum hole and a spectrum opportunity. Therefore, spectrum sensing can be performed in the time, frequency, spatial and code domains.

• Due to recent development of beamforming technology, multiple users can utilize the same channel/frequency at the same time in the same geographical location. Thus, if a primary user does not transmit in all the directions, extra spectrum opportunities can be created for secondary users in the directions where the primary user is not operating, and spectrum sensing needs also to take the angle of arrivals into account.

• Primary users can also use their assigned bands by means of spread spectrum or frequency hopping, and then secondary users can transmit in the same band simultaneously without severely interfering with primary users as long as they adopt an orthogonal code with respect to the codes adopted by primary users. This creates spectrum opportunities in code domain, but meanwhile requires detection of the codes used by primary users as well as multipath parameters.

To perform spectrum sensing, the secondary user performs local measurements of the signal received by the primary users. However the detection of primary user in general is difficult. According to the a priori information required for detection, the resulting complexity and accuracy, spectrum sensing techniques can be categorized in the following types: Energy detector, Feature detector, Matched filtering and coherent detection and Cooperative sensing.

10.4.1) Energy Detector: Energy detection is the most common type of spectrum sensing because it is easy to implement and requires no prior knowledge about the primary signal. Let,

$x(t)$ is the primary user’s signal to be detected at the local receiver of a secondary user,

$n(t)$ is the additive white Gaussian noise,

h is the channel gain from the primary user’s transmitter to the secondary user’s receiver,

$H_0$ is a null hypothesis, meaning there is no primary user present in the band,

$H_1$ means the primary user’s presence, and

$Y(t)$ is the received signal.

Then hypothesis model for detection is given as,

$H_0 ; y(t) = n(t),$

$H_1 ; y(t) = hx(t) + n(t)$

The detection statistics T of energy detector is defined as the average (total) energy of N observed samples,

$T = \frac{1}{N} \sum_{t=1}^N |y(t)|^2$

The decision on whether the spectrum is being occupied by the primary user is made by comparing the detection statistics T with a predetermined threshold $\lambda$. The performance of the detector is characterized by two probabilities: the probability of false alarm ($P_F$ )and the probability of detection ($P_D$).

$P_F$ denotes the probability that the hypothesis test decides the $H_1$ ( presence of primary user) while it is actually $H_0$ (no primary user’s presence) i.e.,

$P_F = P_r(T \gt \lambda : H_0)$

$P_D$ denotes the probability that the test correctly decides $H_1$ i.e.,

$P_D= P_r(T \gt \lambda : H_1)$

A good detector should ensure a high detection probability $P_D$ and a low false alarm probability $P_F$. A well chosen detection threshold can minimize spectrum sensing error, provides the primary user enough protection, and enhances spectrum utilization. Besides its low computational and implementation complexity and short detection time, some challenges also exist in designing a good energy detector. First, the detection threshold depends on the noise power, which may change over time and hence it is difficult to measure precisely in real time. In low signal to noise ratio (SNR) regimes where the noise power is very high, reliable detection of a primary user is even impossible. Moreover, an energy detector can only decide the primary user’s presence by comparing the received signal energy with a threshold; thus, it cannot differentiate the primary user from other unknown signal sources. As such, it can trigger false alarm frequently.

10.4.2) Feature Detector : There are specific features associated with the signal transmitted from a primary user. Features refer to the inherent characteristics of the signal. For example, the transmitted signals are periodic in many communication due to inherent periodicities such as the modulation rate, carrier frequency etc. Such features are usually viewed as the cyclo stationary features, based on which a detector can distinguish cyclo stationary signals from stationary noise. As in most communication systems, the transmitted signals are modulated signals with sine wave carriers, pulse trains, hopping sequences, or cyclic prefixes, while the additive noise is generally wide-sense stationary (WSS) with no correlation. Cyclo stationary feature detectors can be utilized to differentiate noise from primary users’ signal. It is different from an energy detector in the sense that it uses time-domain signal energy as test statistics, while a cyclo stationary feature detector performs a transformation from the time-domain into the frequency feature domain and then conducts a hypothesis test in the new domain. Specifically, the cyclic autocorrelation function (CAF), $ R_y^{\alpha} (\tau)$ of the received signal $y(t)$ is defined as,

$R_y^{\alpha} (\tau) = E [y(t + \tau)y*(t - \tau)e^{j2\pi \alpha t}$

Where $E[.]$ is the expectation operation, $\text{*}$ denotes complex conjugation, and $\alpha$ is the cyclic frequency. Since periodicity is a common property of wireless modulated signals, while noise is Wide Sense Stationary, the CAF of the received signal also demonstrates periodicity when the primary signal is present. Thus, we can represent the CAF using its Fourier series expansion, called the cyclic spectrum density (CSD) function, $S(f,\alpha)$ expressed as,

$S(f,\alpha) = \sum_{\tau= -\infty}^{\infty}R_y^{\alpha}(\tau)e^{-j2 \pi f\tau}$

The CSD function have peaks when the cyclic frequency $\alpha$ equals to the fundamental frequencies of the transmitted signal, i.e., with being the period of $T= \alpha / T_x$ , where $T_x$ is the period of $x(t)$. Here hypothesis is defined as follows,

$H_0$: a null hypothesis, meaning that CSD function does not have any peaks since the noise is non-cyclostationary signals (there is no primary user present in the band).

$H_1$: CSD function has peaks means the primary user’s presence, and

A peak detector can be used to distinguish among the two hypothesis. Different primary communication systems using different air interfaces (modulation, multiplexing, coding, etc.) can also be differentiated by their different properties of cyclostationarity

Compared to energy detectors, that are prone to high false alarm probability due to noise uncertainty and cannot detect weak signals in noise, cyclostationary (feature) detectors are good alternatives because they can differentiate noise from primary users signal and have better detection robustness in low SNR regime.

Generalized Feature Detection: It refers to detection and classification that extracts more feature information other than the cyclostationarity due to the modulated primary signals, such as :

• Transmission technologies used by a primary user,
• The amount of energy and its distribution across different frequencies,
• Channel bandwidth and its shape,
• Power spectrum density,
• Center frequency,
• Idle guard interval of OFDM,
• FFT- type feature etc.

Thus primary user can be identified by extracting the features from the received signal and matching to the a priori information about primary users’ transmission characteristics.

Matched Filtering and Coherent Detection: If secondary users know some information about a primary user’s signal a priori, then the matched filtering is the optimal detection method. A matched filter correlates the already known primary signal with the received signal to detect the presence of the primary user and thus maximizes the SNR in the presence of additive stochastic noise. The main advantage of matched filtering is that it needs less received signal samples and hence requires short time to get low probability of missed detection and false alarm. However, the required number of signal samples also grows as the received SNR decreases, so there exists a SNR wall for a matched filter. In addition, its implementation complexity and power consumption is too high, because the matched filter needs receivers for all types of signals and corresponding receiver algorithms to be executed.

Matched filtering requires perfect knowledge of the primary user’s signal, such as the operating frequency, bandwidth, modulation type and order, pulse shape, packet format, etc.. If wrong information is used for matched filtering, the detection performance will be degraded a lot. On the other hand, most wireless communication systems are used to assist control, equalization, synchronization, continuity, or reference purposes. Even though perfect information of a primary user’s signal may not be attainable, if a certain pattern is known from the received signals, coherent detection ( waveform-based sensing) can be used to decide whether a primary user is transmitting or not. As an example, the procedure of coherent detection using pilot pattern is explained as follows, There are two hypothesis in the coherent detection:

Let

$x_p(t)$ is a known pilot tone,

$\in$ is the fraction of energy allocated to the pilot tone,

$x(t)$ is the desired signal assumed to be orthogonal to the pilot tone,

$n(t)$ is additive white noise,

$H_0$ : Null hypothesis, absence of primary user,

$H_1$: The presence of primary user.

$H_0 : y(t) = n(t)$,

$H_1 : y(t) = \sqrt{\in} x_{p(t)} + \sqrt{1 - \in} x(t) + n(t)$

The test statistics T of the coherent detection is defined as the projected received signal in the pilot direction, i.e.,

$T = \frac{1}{N} \sum_{t=1}^N y(t) \hat x_p (t)$

With $x_p$ is a normalized unit vector in the direction of the pilot tone. As N increases, the test statistics under hypothesis $H_1$ is much greater than that under $H_0$ . By comparing with a predetermined detection threshold, one can decide the presence of a primary user.

Coherent detection is shown to be robust to noise uncertainty, and not limited by the SNR wall as N is large enough. Moreover, coherent detection outperforms energy detection in the sensing convergence time, because the sensing time of energy detection increases quadratically with the SNR reduction, while that of coherent detection only increases linearly. However, information about the waveform patterns is a prerequisite for implementing coherent detection; the more precise information a coherent detector has, the better the sensing performance will be. Table 2 shows the summary of main spectrum sensing techniques.

Table 2: Comparison between various Spectrum Sensing Techniques:

Cooperative Sensing

It is observed that the performance of spectrum sensing is limited by noise uncertainty, shadowing, and multi-path fading effect. When the received primary SNR is too low, there exists a SNR wall, below which reliable spectrum detection is impossible even with a very long sensing time. If secondary users cannot detect the primary transmitter, while the primary receiver is within the secondary users’ transmission range, a hidden primary user problem will occur, and the primary user’s transmission will be interfered.

By taking advantage of the independent fading channels (i.e., spatial diversity) and multiuser diversity, cooperative spectrum sensing is proposed to improve the reliability of spectrum sensing, increase the detection probability to better protect a primary user, and reduce false alarm to utilize the idle spectrum more efficiently. In centralized cooperative spectrum sensing, a central controller, e.g., a secondary Base Station, collects local observations from multiple secondary users, decides the available spectrum channels using some decision fusion rule, and informs the secondary users which channels to access. In distributed cooperative spectrum sensing, secondary users exchange their local detection results among themselves without requiring a backbone infrastructure with reduced cost. Performance of cooperative spectrum sensing depends on the correctness of the local sensing data reported by the secondary users. If malicious users enter a legitimate secondary network and compromise the secondary users, false detection results will be reported to the fusion center, and this kind of attack is called spectrum sensing data falsification (SSDF).