【Advanced Chapter】Signal Spectrum Sensing and Dynamic Spectrum Allocation in MATLAB

# 2.1 Basic Principles of Spectrum Sensing

### 2.1.1 Definition and Purpose of Spectrum Sensing

Spectrum sensing refers to the process of detecting and analyzing the usage of the radio frequency spectrum within a given time and spatial range. Its purpose is to understand the occupancy of the spectrum, identify idle frequency bands, and provide the best spectral resource allocation for wireless communication systems.

### 2.1.2 Classification and Methods of Spectrum Sensing

According to the sensing target, spectrum sensing can be classified into the following two categories:

- **Spectrum Occupancy Sensing:** Detecting and identifying occupied frequency bands to avoid interference with existing users.
- **Spectrum Idle Sensing:** Finding and identifying unoccupied frequency bands to provide access opportunities for new users.

Various methods are used for spectrum sensing, including:

- Energy Detection Algorithms
- Cyclostationary Feature Detection Algorithms
- Feature Extraction Algorithms

# 2. Spectrum Sensing Algorithms and Techniques

### 2.1 Basic Principles of Spectrum Sensing

#### 2.1.1 Definition and Purpose

Spectrum sensing refers to the process of detecting and identifying unauthorized or unused spectrum resources within a given frequency band range. Its purpose is to provide cognitive radio systems with perceptual information about available spectrum without interfering with authorized users' normal communications.

#### 2.1.2 Classification and Methods

Spectrum sensing algorithms can be classified into the following two categories:

- **Non-Cooperative Spectrum Sensing:** Relying solely on the signals received by oneself, without depending on feedback from authorized users, for spectrum sensing.
- **Cooperative Spectrum Sensing:** Uti***

***mon spectrum sensing methods include:

- Energy Detection Algorithms
- Cyclostationary Feature Detection Algorithms
- Feature Extraction Algorithms

### 2.2 Spectrum Sensing Algorithms

#### 2.2.1 Energy Detection Algorithm

The energy detection algorithm is the simplest spectrum sensing algorithm. It compares the energy of the received signal with a given threshold. If the energy of the received signal exceeds the threshold, the frequency band is considered occupied.

% Input signal
signal = randn(1, 1000);

% Threshold
threshold = 0.5;

% Energy detection
energy = sum(abs(signal).^2);

% Determine the occupancy of the frequency band
if energy > threshold
    disp('Frequency band is occupied');
else
    disp('Frequency band is not occupied');
end

**Code Logic Analysis:**

- `randn(1, 1000)`: Generate 1000 random Gaussian distribution data points, simulating the received signal.
- `sum(abs(signal).^2)`: Calculate the signal's energy, i.e., the sum of the squares of the signal amplitude.
- `threshold`: Set the energy detection threshold.
- `if energy > threshold`: If the signal energy exceeds the threshold, the frequency band is considered occupied.
- `else`: Otherwise, the frequency band is considered not occupied.

#### 2.2.2 Cyclostationary Feature Detection Algorithm

The cyclostationary feature detection algorithm detects the presence of cyclostationary signals in the spectrum by calculating the cyclostationary feature density of the received signal. If cyclostationary signals are detected, the frequency band is considered occupied.

% Input signal
signal = sin(2 * pi * 10 * (0:0.01:1));

% Cyclostationary feature density calculation
psd = pwelch(signal, [], [], [], 100);

% Spectrum plot
figure;
plot(psd);
xlabel('Frequency (Hz)');
ylabel('Power Spectral Density (dB/Hz)');

% Frequency band occupancy determination
if max(psd) > -50
    disp('Frequency band is occupied');
else
    disp('Frequency band is not occupied');
end

**Code Logic Analysis:**

- `sin(2 * pi * 10 * (0:0.01:1))`: Generate a 10 Hz sine wave signal, simulating the received signal.
- `pwelch`: Calculate the cyclostationary feature density of the signal.
- `plot(psd)`: Plot the cyclostationary feature density graph.
- `max(psd) > -50`: Set the cyclostationary feature density threshold. If the maximum value is greater than the threshold, the frequency band is considered occupied.
- `else`: Otherwise, the frequency band is considered not occupied.

#### 2.2.3 Feature Extraction Algorithm

The feature extraction algorithm identi***mon feature parameters include peak value, center frequency, bandwidth, etc.

% Input signal
signal = randn(1, 1000);

% Feature extraction
features = [max(signal), mean(signal), std(signal)];

% Feature classification
classifier = fitcsvm(features, ones(1, length(signal)));

% Frequency band occupancy determination
if predict(classifier, features) == 1
    disp('Frequency band is occupied');
else
    disp('Frequency band is not occupied');
end

**Code Logic Analysis:**

- `randn(1, 1000)`: Generate 1000 random Gaussian distribution data points, simulating the received signal.
- `[max(signal), mean(signal), std(signal)]`: Extract the peak value, mean, and standard deviation of the signal as feature parameters.
- `fitcsvm`: Train a support vector machine classifier to map feature parameters to the frequency band occupancy state.
- `predict(classifier, features)`: Use the classifier to predict the frequency band occupancy state.
- `if predict(classifier, features) == 1`: If the classifier predicts that the frequency band is occupied, then it is determined to be occupied.
- `else`: Otherwise, it is determined not to be occupied.

# 3. Dynamic Spectrum Allocation Mechanism

### 3.1 Basic Concepts of Dynamic Spectrum Allocation

#### 3.1.1 Significance and Advantages of Dynamic Spectrum Allocation

Dynamic Spectrum Allocation (DSA) is a radio spectrum management technology that allows authorized and unauthorized users to dynamically share spectrum resources without interfering with existing users. The primary significance and advantages of DSA include:

- **Increased Spectrum Utilization:** DSA allows the allocation of spectrum based on demand at different times and locations, thereby increasing spectrum utilization.
- **Spectrum Flexibility:** DSA allows users to access and release spectrum flexibly based on need, thereby increasing spectrum flexibility.
- **Reduced Interference:** DSA reduces interference between different users by coordinating spectrum allocation.
- **Promotion of Innovation:**