Eric Rebeiz

Energy Detection Based Spectrum Sensing Over

Energy detection (ED) is a simple and popular method of spectrum sensing in cognitive radio systems. It is also widely known that the performance of sensing techniques is largely affected when users experience fading effects. This paper investigates the performance of an energy detector over generalized κ-μ and κ- μ extreme fading channels, which have been shown to provide remarkably accurate fading characterization. Novel analytic expressions are firstly derived for the corresponding average probability of detection for the case of single-user detection. These results are subsequently extended to the case of square-law selection (SLS) diversity and for collaborative detection scenarios. As expected, the performance of the detector is highly dependent upon the severity of fading since even small variations of the fading conditions affect significantly the value of the average probability of detection. Furthermore, the performance of the detector improves substantially as the number of branches or collaborating users increase in both severe and moderate fading conditions, whereas it is shown that the κ- μ extreme model is capable of accounting for fading variations even at low signal-to-noise values. The offered results are particularly useful in assessing the effect of fading in ED-based cognitive radio communication systems; therefore, they can be used in quantifying the associated tradeoffs between sensing performance and energy efficiency in cognitive radio networks.

\kappa{-}\mu

In the context of Cognitive Radios (CRs), cyclostationary detection of primary users (PUs) is regarded as a common method for spectrum sensing. Cyclostationary detectors rely on the knowledge of the signals symbol rate, carrier frequency, and modulation class in order to detect the present cyclic features. Cyclic frequency and sampling clock offsets are the two receiver impairments considered in this work. Cyclic frequency offsets, which occur due to oscillator frequency offsets, Doppler shifts, or imperfect knowledge of the cyclic frequencies, result in computing the test statistic at an offset from the true cyclic frequency. In this paper, we analyze the effect of cyclic frequency offsets on conventional cyclostationary detection, and propose a new multi-frame test statistic that reduces the degradation due to cyclic frequency offsets. Due to the multi-frame processing of the proposed statistic, non-coherent integration might occur across frames. Through an optimization framework developed in this work that can be performed offline, we determine the best frame length that maximizes the average detection performance of the proposed cyclostationary detection method given the statistical distributions of the receiver impairments. As a result of the optimization, the proposed detectors is shown to achieve the performance gains over conventional detectors given the constrained sensing time. We derive the proposed detectors theoretical average detection performance, and compare it to the performance of the conventional cyclostationary detector. Our analysis shows that gains in average detection performance using the proposed method can be achieved when the effect of sampling clock offset is less severe than that of the cyclic frequency offset. The analysis given in this paper can be used as a design guideline for practical implementation of cyclostationary spectrum sensors.

and

In this paper, we propose and analyze a spectrum sensing method based on cyclostationarity specifically targeted for receivers with multiple antennas. This detection method is used for determining the presence or absence of primary users in cognitive radio networks based on the eigenvalues of the cyclic covariance matrix of received signals. In particular, the cyclic correlation significance test is used to detect a specific signal-of-interest by exploiting knowledge of its cyclic frequencies. Analytical asymptotic expressions for the probability of detection and probability of false-alarm under both spatially uncorrelated and spatially correlated noise are derived and verified by simulations. The detection performance in a Rayleigh flat-fading environment is found and verified through simulations. One of the advantages of the proposed method is that the detection threshold is shown to be independent of both the number of samples and the noise covariance, effectively eliminating the dependence on accurate noise estimation. The proposed method is also shown to provide higher detection probability and better robustness to noise uncertainty than existing multiple antenna cyclostationary-based spectrum sensing algorithms under both AWGN as well as a quasi-static Rayleigh fading channel.

\kappa{-}\mu

We present a novel modulation level classification (MLC) method based on probability distribution distance functions. The proposed method uses modified Kuiper and Kolmogorov-Smirnov distances to achieve low computational complexity and outperforms the state of the art methods based on cumulants and goodness-of-fit tests. We derive the theoretical performance of the proposed MLC method and verify it via simulations. The best classification accuracy, under AWGN with SNR mismatch and phase jitter, is achieved with the proposed MLC method using Kuiper distances.

Extreme Fading Channels

Detecting the presence of licensed users and avoiding interference to them is vital to the proper operation of a Cognitive Radio (CR) network. Operating in a wideband channel requires high Nyquist sampling rates, which is limited by the state-of-the-art A/D converters. Compressive sampling is a promising solution to reduce sampling rates required in modern wideband communication systems. Among various signal detectors, feature detectors which exploit a signal cyclostationarity are robust against noise uncertainties. In this paper, we exploit the sparsity of the two-dimensional spectral correlation function (SCF), and propose a reduced complexity reconstruction method of the Nyquist SCF from the sub-Nyquist samples. The reconstruction optimization is formulated as a regularized least squares problem, and its closed form solution is derived. We show that for a given spectrum sparsity, there exists a lower bound on sampling rates that allows reliable SCF reconstruction.

Optimizing Wideband Cyclostationary Spectrum Sensing Under Receiver Impairments

Blind modulation classification is of vital importance in spectrum surveillance applications and future heterogeneous wireless networks. In standardized wireless systems, modulation classification can be performed through exhaustive search of known signal features. Most commonly used classifiers are based on the detection of cyclostationary features, which are second-order moments of a signal, related to its carrier and symbol rate. However, when the signal parameters are unknown, an exhaustive search for cyclostationary features is energy inefficient due to high computational complexity. In this paper, we present a reconfigurable processor architecture that can blindly classify any linearly modulated signal (M-QAM, M-PSK, M-ASK, and GMSK) in addition to multi-carrier signals and spread spectrum signals. The contributions of this work are twofold. First, we analyze the complexity tradeoffs among different dependent signal processing kernels in order to minimize the total processing time and energy. Second, we optimize the processor architecture by the co-design methodology to enhance block reusability and reconfigurability. The proposed processor has been verified and synthesized in a 40-nm CMOS technology with core area of 0.06 mm2 and power dissipation of 10 mW under 0.9 V supply voltage at 500 MHz. Under a 500 MHz wide-band signal at 10 dB SNR, a complete blind classification process consumes 10.37 μJ to meet 95% of classification accuracy.

Multiple Antenna Cyclostationary Spectrum Sensing Based on the Cyclic Correlation Significance Test

Wideband spectrum sensing which requires detecting the presence or absence of signals in a wideband channel faces multiple practical issues. Current bandwidth limitations of state-of-the-art analog to digital converters require alternative approaches to be considered for wideband sensing. Cyclostationary feature detection is a promising sensing tool which is robust to noise, and takes advantage of the noise stationarity. In this paper, we propose a cyclostationary feature detector that operates on sub-Nyquist samples obtained via either multicoset sampling or the modulated wideband converter analog front-end, and present the receivers detection performance at low SNRs.

Computationally Efficient Modulation Level Classification Based on Probability Distribution Distance Functions

In this paper, we propose a reduced-complexity modulation classifier using multi-cycle features extracted from the Spectral Correlation Function (SCF) in order to distinguish among QAM, BPSK, MSK and AM modulation schemes. We analytically derive SCF statistics of the noise and signal features used for classification for finite number of samples, and use Chebyshev inequality to upper bound the minimum number of spectral averages required to attain a predetermined correct classification probability. Both theoretical and simulation results show that the proposed classifier requires on the order of 50 spectral averages to achieve a correct classification probability of 0.9 at SNR = 5 dB. The algorithm and corresponding analysis presented in this paper can be extended to classify other modulation schemes.

Cyclostationary-based low complexity wideband spectrum sensing using compressive sampling

In the context of Cognitive Radio (CR), opportunistic transmissions can exploit temporarily vacant spectral bands. Efficient and reliable spectrum sensing is a key in the CR process. CR receivers traditionally deal with wideband signals with high Nyquist rates and low Signal to Noise Ratios (SNRs). Thus, in this paper, we propose sub-Nyquist sampling and cyclostationary detection, which is robust to noise. We first reconstruct the cyclic spectrum or Spectral Correlation Function (SCF) of the signal, which is a characteristic function of cyclostationary signals such as communication signals, from sub-Nyquist samples and then perform detection. We consider both sparse and non sparse signals as well as blind and non blind detection in the sparse case. For each one of those scenarii, we derive the minimal sampling rate allowing for perfect reconstruction of the signals SCF in a noise-free environment and provide SCF recovery techniques. In the simulations, we show SCF recovery at the minimal rate in noise-free settings as well as the performance of our detector in the presence of noise.

Energy-Efficient Processor for Blind Signal Classification in Cognitive Radio Networks

Cyclostationary detection involves detecting cyclic features of modulated signals which are functions of various transmit parameters including the symbol rate, carrier frequency, and modulation format. However, imperfect knowledge of these transmit parameters at the sensing radio results in computing the detection test statistic under a cyclic frequency offset (CFO). The detection performance of the conventional cyclic autocorrelation function has been shown to degrade in the presence of the CFO impairment. In this paper, we propose a new multi-frame detection statistic that improves the robustness of the conventional cyclostationary detector to CFO impairments. The achievable gains in using this method over conventional detectors are quantified analytically and verified through hardware experiments using the Universal Software Radio Platform (USRP N200) transceivers.