Open Access

A Review on Spectrum Sensing for Cognitive Radio: Challenges and Solutions

  • Yonghong Zeng1Email author,
  • Ying-Chang Liang1,
  • Anh Tuan Hoang1 and
  • Rui Zhang1
EURASIP Journal on Advances in Signal Processing20102010:381465

DOI: 10.1155/2010/381465

Received: 13 May 2009

Accepted: 9 October 2009

Published: 6 January 2010

Abstract

Cognitive radio is widely expected to be the next Big Bang in wireless communications. Spectrum sensing, that is, detecting the presence of the primary users in a licensed spectrum, is a fundamental problem for cognitive radio. As a result, spectrum sensing has reborn as a very active research area in recent years despite its long history. In this paper, spectrum sensing techniques from the optimal likelihood ratio test to energy detection, matched filtering detection, cyclostationary detection, eigenvalue-based sensing, joint space-time sensing, and robust sensing methods are reviewed. Cooperative spectrum sensing with multiple receivers is also discussed. Special attention is paid to sensing methods that need little prior information on the source signal and the propagation channel. Practical challenges such as noise power uncertainty are discussed and possible solutions are provided. Theoretical analysis on the test statistic distribution and threshold setting is also investigated.

1. Introduction

It was shown in a recent report [1] by the USA Federal Communications Commission (FCC) that the conventional fixed spectrum allocation rules have resulted in low spectrum usage efficiency in almost all currently deployed frequency bands. Measurements in other countries also have shown similar results [2]. Cognitive radio, first proposed in [3], is a promising technology to fully exploit the under-utilized spectrum, and consequently it is now widely expected to be the next Big Bang in wireless communications. There have been tremendous academic researches on cognitive radios, for example, [4, 5], as well as application initiatives, such as the IEEE 802.22 standard on wireless regional area network (WRAN) [6, 7] and the Wireless Innovation Alliance [8] including Google and Microsoft as members, which advocate to unlock the potential in the so-called "White Spaces" in the television (TV) spectrum. The basic idea of a cognitive radio is spectral reusing or spectrum sharing, which allows the secondary networks/users to communicate over the spectrum allocated/licensed to the primary users when they are not fully utilizing it. To do so, the secondary users are required to frequently perform spectrum sensing, that is, detecting the presence of the primary users. Whenever the primary users become active, the secondary users have to detect the presence of them with a high probability and vacate the channel or reduce transmit power within certain amount of time. For example, for the upcoming IEEE 802.22 standard, it is required for the secondary users to detect the TV and wireless microphone signals and vacant the channel within two seconds once they become active. Furthermore, for TV signal detection, it is required to achieve 90% probability of detection and 10% probability of false alarm at signal-to-noise ratio (SNR) level as low as https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq1_HTML.gif  dB.

There are several factors that make spectrum sensing practically challenging. First, the required SNR for detection may be very low. For example, even if a primary transmitter is near a secondary user (the detection node), the transmitted signal of the primary user can be deep faded such that the primary signal's SNR at the secondary receiver is well below https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq2_HTML.gif  dB. However, the secondary user still needs to detect the primary user and avoid using the channel because it may strongly interfere with the primary receiver if it transmits. A practical scenario of this is a wireless microphone operating in TV bands, which only transmits with a power less than 50 mW and a bandwidth less than 200 KHz. If a secondary user is several hundred meters away from the microphone device, the received SNR may be well below https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq3_HTML.gif  dB. Secondly, multipath fading and time dispersion of the wireless channels complicate the sensing problem. Multipath fading may cause the signal power to fluctuate as much as https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq4_HTML.gif  dB. On the other hand, unknown time dispersion in wireless channels may turn the coherent detection unreliable. Thirdly, the noise/interference level may change with time and location, which yields the noise power uncertainty issue for detection [912].

Facing these challenges, spectrum sensing has reborn as a very active research area over recent years despite its long history. Quite a few sensing methods have been proposed, including the classic likelihood ratio test (LRT) [13], energy detection (ED) [9, 10, 13, 14], matched filtering (MF) detection [10, 13, 15], cyclostationary detection (CSD) [1619], and some newly emerging methods such as eigenvalue-based sensing [6, 2025], wavelet-based sensing [26], covariance-based sensing [6, 27, 28], and blindly combined energy detection [29]. These methods have different requirements for implementation and accordingly can be classified into three general categories: (a) methods requiring both source signal and noise power information, (b) methods requiring only noise power information (semiblind detection), and (c) methods requiring no information on source signal or noise power (totally blind detection). For example, LRT, MF, and CSD belong to category A; ED and wavelet-based sensing methods belong to category B; eigenvalue-based sensing, covariance-based sensing, and blindly combined energy detection belong to category C. In this paper, we focus on methods in categories B and C, although some other methods in category A are also discussed for the sake of completeness. Multiantenna/receiver systems have been widely deployed to increase the channel capacity or improve the transmission reliability in wireless communications. In addition, multiple antennas/receivers are commonly used to form an array radar [30, 31] or a multiple-input multiple-output (MIMO) radar [32, 33] to enhance the performance of range, direction, and/or velocity estimations. Consequently, MIMO techniques can also be applied to improve the performance of spectrum sensing. Therefore, in this paper we assume a multi-antenna system model in general, while the single-antenna system is treated as a special case.

When there are multiple secondary users/receivers distributed at different locations, it is possible for them to cooperate to achieve higher sensing reliability. There are various sensing cooperation schemes in the current literature [3444]. In general, these schemes can be classified into two categories: (A) data fusion: each user sends its raw data or processed data to a specific user, which processes the data collected and then makes the final decision; (B) decision fusion: multiple users process their data independently and send their decisions to a specific user, which then makes the final decision.

In this paper, we will review various spectrum sensing methods from the optimal LRT to practical joint space-time sensing, robust sensing, and cooperative sensing and discuss their advantages and disadvantages. We will pay special attention to sensing methods with practical application potentials. The focus of this paper is on practical sensing algorithm designs; for other aspects of spectrum sensing in cognitive radio, the interested readers may refer to other resources like [4552].

The rest of this paper is organized as follows. The system model for the general setup with multiple receivers for sensing is given in Section 2. The optimal LRT-based sensing due to the Neyman-Pearson theorem is reviewed in Section 3. Under some special conditions, it is shown that the LRT becomes equivalent to the estimator-correlator detection, energy detection, or matched filtering detection. The Bayesian method and the generalized LRT for sensing are discussed in Section 4. Detection methods based on the spatial correlations among multiple received signals are discussed in Section 5, where optimally combined energy detection and blindly combined energy detection are shown to be optimal under certain conditions. Detection methods combining both spatial and time correlations are reviewed in Section 6, where the eigenvalue-based and covariance-based detections are discussed in particular. The cyclostationary detection, which exploits the statistical features of the primary signals, is reviewed in Section 7. Cooperative sensing is discussed in Section 8. The impacts of noise uncertainty and noise power estimation to the sensing performance are analyzed in Section 9. The test statistic distribution and threshold setting for sensing are reviewed in Section 10, where it is shown that the random matrix theory is very useful for the related study. The robust spectrum sensing to deal with uncertainties in source signal and/or noise power knowledge is reviewed in Section 11, with special emphasis on the robust versions of LRT and matched filtering detection methods. Practical challenges and future research directions for spectrum sensing are discussed in Section 12. Finally, Section 13 concludes the paper.

2. System Model

We assume that there are https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq5_HTML.gif antennas at the receiver. These antennas can be sufficiently close to each other to form an antenna array or well separated from each other. We assume that a centralized unit is available to process the signals from all the antennas. The model under consideration is also applicable to the multinode cooperative sensing [3444, 53], if all nodes are able to send their observed signals to a central node for processing. There are two hypotheses: https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq6_HTML.gif , signal absent, and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq7_HTML.gif , signal present. The received signal at antenna/receiver https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq8_HTML.gif is given by

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ1_HTML.gif
(1)
In hypothesis https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq9_HTML.gif , https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq10_HTML.gif is the received source signal at antenna/receiver https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq11_HTML.gif , which may include the channel multipath and fading effects. In general, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq12_HTML.gif can be expressed as
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ2_HTML.gif
(2)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq13_HTML.gif denotes the number of primary user/antenna signals, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq14_HTML.gif denotes the transmitted signal from primary user/antenna https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq15_HTML.gif , https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq16_HTML.gif denotes the propagation channel coefficient from the https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq17_HTML.gif th primary user/antenna to the https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq18_HTML.gif th receiver antenna, and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq19_HTML.gif denotes the channel order for https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq20_HTML.gif . It is assumed that the noise samples https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq21_HTML.gif 's are independent and identically distributed (i.i.d) over both https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq22_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq23_HTML.gif . For simplicity, we assume that the signal, noise, and channel coefficients are all real numbers.

The objective of spectrum sensing is to make a decision on the binary hypothesis testing (choose https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq24_HTML.gif or https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq25_HTML.gif ) based on the received signal. If the decision is https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq26_HTML.gif , further information such as signal waveform and modulation schemes may be classified for some applications. However, in this paper, we focus on the basic binary hypothesis testing problem. The performance of a sensing algorithm is generally indicated by two metrics: probability of detection, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq27_HTML.gif , which defines, at the hypothesis https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq28_HTML.gif , the probability of the algorithm correctly detecting the presence of the primary signal; and probability of false alarm, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq29_HTML.gif , which defines, at the hypothesis https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq30_HTML.gif , the probability of the algorithm mistakenly declaring the presence of the primary signal. A sensing algorithm is called "optimal" if it achieves the highest https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq31_HTML.gif for a given https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq32_HTML.gif with a fixed number of samples, though there could be other criteria to evaluate the performance of a sensing algorithm.

Stacking the signals from the https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq33_HTML.gif antennas/receivers yields the following https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq34_HTML.gif vectors:
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ3_HTML.gif
(3)

The hypothesis testing problem based on https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq35_HTML.gif signal samples is then obtained as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ4_HTML.gif
(4)

3. Neyman-Pearson Theorem

The Neyman-Pearson (NP) theorem [13, 54, 55] states that, for a given probability of false alarm, the test statistic that maximizes the probability of detection is the likelihood ratio test (LRT) defined as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ5_HTML.gif
(5)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq36_HTML.gif denotes the probability density function (PDF), and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq37_HTML.gif denotes the received signal vector that is the aggregation of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq38_HTML.gif Such a likelihood ratio test decides https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq39_HTML.gif when https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq40_HTML.gif exceeds a threshold https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq41_HTML.gif , and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq42_HTML.gif otherwise.

The major difficulty in using the LRT is its requirements on the exact distributions given in (5). Obviously, the distribution of random vector https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq43_HTML.gif under https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq44_HTML.gif is related to the source signal distribution, the wireless channels, and the noise distribution, while the distribution of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq45_HTML.gif under https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq46_HTML.gif is related to the noise distribution. In order to use the LRT, we need to obtain the knowledge of the channels as well as the signal and noise distributions, which is practically difficult to realize.

If we assume that the channels are flat-fading, and the received source signal sample https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq47_HTML.gif 's are independent over https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq48_HTML.gif , the PDFs in LRT are decoupled as
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ6_HTML.gif
(6)
If we further assume that noise and signal samples are both Gaussian distributed, that is, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq49_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq50_HTML.gif , the LRT becomes the estimator-correlator (EC) [13] detector for which the test statistic is given by
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ7_HTML.gif
(7)

From (4), we see that https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq51_HTML.gif is actually the minimum-mean-squared-error (MMSE) estimation of the source signal https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq52_HTML.gif . Thus, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq53_HTML.gif in (7) can be seen as the correlation of the observed signal https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq54_HTML.gif with the MMSE estimation of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq55_HTML.gif .

The EC detector needs to know the source signal covariance matrix https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq56_HTML.gif and noise power https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq57_HTML.gif . When the signal presence is unknown yet, it is unrealistic to require the source signal covariance matrix (related to unknown channels) for detection. Thus, if we further assume that https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq58_HTML.gif , the EC detector in (7) reduces to the well-known energy detector (ED) [9, 14] for which the test statistic is given as follows (by discarding irrelevant constant terms):
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ8_HTML.gif
(8)

Note that for the multi-antenna/receiver case, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq59_HTML.gif is actually the summation of signals from all antennas, which is a straightforward cooperative sensing scheme [41, 56, 57]. In general, the ED is not optimal if https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq60_HTML.gif is non-diagonal.

If we assume that noise is Gaussian distributed and source signal https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq61_HTML.gif is deterministic and known to the receiver, which is the case for radar signal processing [32, 33, 58], it is easy to show that the LRT in this case becomes the matched filtering-based detector, for which the test statistic is
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ9_HTML.gif
(9)

4. Bayesian Method and the Generalized Likelihood Ratio Test

In most practical scenarios, it is impossible to know the likelihood functions exactly, because of the existence of uncertainty about one or more parameters in these functions. For instance, we may not know the noise power https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq62_HTML.gif and/or source signal covariance https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq63_HTML.gif . Hypothesis testing in the presence of uncertain parameters is known as "composite" hypothesis testing. In classic detection theory, there are two main approaches to tackle this problem: the Bayesian method and the generalized likelihood ratio test (GLRT).

In the Bayesian method [13], the objective is to evaluate the likelihood functions needed in the LRT through marginalization, that is,

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ10_HTML.gif
(10)
where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq64_HTML.gif represents all the unknowns when https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq65_HTML.gif is true. Note that the integration operation in (10) should be replaced with a summation if the elements in https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq66_HTML.gif are drawn from a discrete sample space. Critically, we have to assign a prior distribution https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq67_HTML.gif to the unknown parameters. In other words, we need to treat these unknowns as random variables and use their known distributions to express our belief in their values. Similarly, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq68_HTML.gif can be defined. The main drawbacks of the Bayesian approach are listed as follows.
  1. (1)

    The marginalization operation in (10) is often not tractable except for very simple cases.

     
  2. (2)

    The choice of prior distributions affects the detection performance dramatically and thus it is not a trivial task to choose them.

     

To make the LRT applicable, we may estimate the unknown parameters first and then use the estimated parameters in the LRT. Known estimation techniques could be used for this purpose [59]. However, there is one major difference from the conventional estimation problem where we know that signal is present, while in the case of spectrum sensing we are not sure whether there is source signal or not (the first priority here is the detection of signal presence). At different hypothesis ( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq69_HTML.gif or https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq70_HTML.gif ), the unknown parameters are also different.

The GLRT is one efficient method [13, 55] to resolve the above problem, which has been used in many applications, for example, radar and sonar signal processing. For this method, the maximum likelihood (ML) estimation of the unknown parameters under https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq71_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq72_HTML.gif is first obtained as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ11_HTML.gif
(11)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq73_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq74_HTML.gif are the set of unknown parameters under https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq75_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq76_HTML.gif , respectively. Then, the GLRT statistic is formed as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ12_HTML.gif
(12)

Finally, the GLRT decides https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq77_HTML.gif if https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq78_HTML.gif , where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq79_HTML.gif is a threshold, and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq80_HTML.gif otherwise.

It is not guaranteed that the GLRT is optimal or approaches to be optimal when the sample size goes to infinity. Since the unknown parameters in https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq81_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq82_HTML.gif are highly dependent on the noise and signal statistical models, the estimations of them could be vulnerable to the modeling errors. Under the assumption of Gaussian distributed source signals and noises, and flat-fading channels, some efficient spectrum sensing methods based on the GLRT can be found in [60].

5. Exploiting Spatial Correlation of Multiple Received Signals

The received signal samples at different antennas/receivers are usually correlated, because all https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq83_HTML.gif 's are generated from the same source signal https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq84_HTML.gif 's. As mentioned previously, the energy detection defined in (8) is not optimal for this case. Furthermore, it is difficult to realize the LRT in practice. Hence, we consider suboptimal sensing methods as follows.

If https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq85_HTML.gif , and assuming that the propagation channels are flat-fading ( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq86_HTML.gif ) and known to the receiver, the energy at different antennas can be coherently combined to obtain a nearly optimal detection [41, 43, 57]. This is also called maximum ratio combining (MRC). However, in practice, the channel coefficients are unknown at the receiver. As a result, the coherent combining may not be applicable and the equal gain combining (EGC) is used in practice [41, 57], which is the same as the energy detection defined in (8).

In general, we can choose a matrix https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq87_HTML.gif with https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq88_HTML.gif rows to combine the signals from all antennas as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ13_HTML.gif
(13)

The combining matrix should be chosen such that the resultant signal has the largest SNR. It is obvious that the SNR after combining is

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ14_HTML.gif
(14)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq89_HTML.gif denotes the mathematical expectation. Hence, the optimal combining matrix should maximize the value of function https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq90_HTML.gif . Let https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq91_HTML.gif be the statistical covariance matrix of the primary signals. It can be verified that

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ15_HTML.gif
(15)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq92_HTML.gif denotes the trace of a matrix. Let https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq93_HTML.gif be the maximum eigenvalue of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq94_HTML.gif and let https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq95_HTML.gif be the corresponding eigenvector. It can be proved that the optimal combining matrix degrades to the vector https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq96_HTML.gif [29].

Upon substituting https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq97_HTML.gif into (13), the test statistic for the energy detection becomes
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ16_HTML.gif
(16)

The resulting detection method is called optimally combined energy detection (OCED) [29]. It is easy to show that this test statistic is better than https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq98_HTML.gif in terms of SNR.

The OCED needs an eigenvector of the received source signal covariance matrix, which is usually unknown. To overcome this difficulty, we provide a method to estimate the eigenvector using the received signal samples only. Considering the statistical covariance matrix of the signal defined as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ17_HTML.gif
(17)

we can verify that

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ18_HTML.gif
(18)
Since https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq99_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq100_HTML.gif have the same eigenvectors, the vector https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq101_HTML.gif is also the eigenvector of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq102_HTML.gif corresponding to its maximum eigenvalue. However, in practice, we do not know the statistical covariance matrix https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq103_HTML.gif either, and therefore we cannot obtain the exact vector https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq104_HTML.gif . An approximation of the statistical covariance matrix is the sample covariance matrix defined as
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ19_HTML.gif
(19)

Let https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq105_HTML.gif (normalized to https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq106_HTML.gif ) be the eigenvector of the sample covariance matrix corresponding to its maximum eigenvalue. We can replace the combining vector https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq107_HTML.gif by https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq108_HTML.gif , that is,

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ20_HTML.gif
(20)

Then, the test statistics for the resulting blindly combined energy detection (BCED) [29] becomes

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ21_HTML.gif
(21)

It can be verified that

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ22_HTML.gif
(22)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq109_HTML.gif is the maximum eigenvalue of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq110_HTML.gif . Thus, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq111_HTML.gif can be taken as the maximum eigenvalue of the sample covariance matrix. Note that this test is a special case of the eigenvalue-based detection (EBD) [2025].

6. Combining Space and Time Correlation

In addition to being spatially correlated, the received signal samples are usually correlated in time due to the following reasons.

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq112_HTML.gif The received signal is oversampled. Let https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq113_HTML.gif be the Nyquist sampling period of continuous-time signal https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq114_HTML.gif and let https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq115_HTML.gif be the sampled signal based on the Nyquist sampling rate. Thanks to the Nyquist theorem, the signal https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq116_HTML.gif can be expressed as
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ23_HTML.gif
(23)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq117_HTML.gif is an interpolation function. Hence, the signal samples https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq118_HTML.gif are only related to https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq119_HTML.gif , where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq120_HTML.gif is the actual sampling period. If the sampling rate at the receiver is https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq121_HTML.gif , that is, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq122_HTML.gif , then https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq123_HTML.gif must be correlated over https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq124_HTML.gif . An example of this is the wireless microphone signal specified in the IEEE 802.22 standard [6, 7], which occupies about 200 KHz in a 6-MHz TV band. In this example, if we sample the received signal with sampling rate no lower than 6 MHz, the wireless microphone signal is actually oversampled and the resulting signal samples are highly correlated in time.

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq125_HTML.gif The propagation channel is time-dispersive. In this case, the received signal can be expressed as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ24_HTML.gif
(24)
where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq126_HTML.gif is the transmitted signal and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq127_HTML.gif is the response of the time-dispersive channel. Since the sampling period https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq128_HTML.gif is usually very small, the integration (24) can be approximated as
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ25_HTML.gif
(25)
Hence,
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ26_HTML.gif
(26)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq129_HTML.gif is the support of the channel response https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq130_HTML.gif , with https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq131_HTML.gif for https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq132_HTML.gif . For time-dispersive channels, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq133_HTML.gif and thus even if the original signal samples https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq134_HTML.gif 's are i.i.d., the received signal samples https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq135_HTML.gif 's are correlated.

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq136_HTML.gif The transmitted signal is correlated in time. In this case, even if the channel is flat-fading and there is no oversampling at the receiver, the received signal samples are correlated.

The above discussions suggest that the assumption of independent (in time) received signal samples may be invalid in practice, such that the detection methods relying on this assumption may not perform optimally. However, additional correlation in time may not be harmful for signal detection, while the problem is how we can exploit this property. For the multi-antenna/receiver case, the received signal samples are also correlated in space. Thus, to use both the space and time correlations, we may stack the signals from the https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq137_HTML.gif antennas and over https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq138_HTML.gif sampling periods all together and define the corresponding https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq139_HTML.gif signal/noise vectors:
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ27_HTML.gif
(27)
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ28_HTML.gif
(28)
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ29_HTML.gif
(29)

Then, by replacing https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq140_HTML.gif by https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq141_HTML.gif , we can directly extend the previously introduced OCED and BCED methods to incorporate joint space-time processing. Similarly, the eigenvalue-based detection methods [2124] can also be modified to work for correlated signals in both time and space. Another approach to make use of space-time signal correlation is the covariance based detection [27, 28, 61] briefly described as follows. Defining the space-time statistical covariance matrices for the signal and noise as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ30_HTML.gif
(30)

respectively, we can verify that

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ31_HTML.gif
(31)

If the signal is not present, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq142_HTML.gif , and thus the off-diagonal elements in https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq143_HTML.gif are all zeros. If there is a signal and the signal samples are correlated, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq144_HTML.gif is not a diagonal matrix. Hence, the nonzero off-diagonal elements of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq145_HTML.gif can be used for signal detection.

In practice, the statistical covariance matrix can only be computed using a limited number of signal samples, where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq146_HTML.gif can be approximated by the sample covariance matrix defined as
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ32_HTML.gif
(32)
Based on the sample covariance matrix, we could develop the covariance absolute value (CAV) test [27, 28] defined as
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ33_HTML.gif
(33)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq147_HTML.gif denotes the https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq148_HTML.gif th element of the sample covariance matrix https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq149_HTML.gif .

There are other ways to utilize the elements in the sample covariance matrix, for example, the maximum value of the nondiagonal elements, to form different test statistics. Especially, when we have some prior information on the source signal correlation, we may choose a corresponding subset of the elements in the sample covariance matrix to form a more efficient test.

Another effective usage of the covariance matrix for sensing is the eigenvalue based detection (EBD) [2025], which uses the eigenvalues of the covariance matrix as test statistics.

7. Cyclostationary Detection

Practical communication signals may have special statistical features. For example, digital modulated signals have nonrandom components such as double sidedness due to sinewave carrier and keying rate due to symbol period. Such signals have a special statistical feature called cyclostationarity, that is, their statistical parameters vary periodically in time. This cyclostationarity can be extracted by the spectral-correlation density (SCD) function [1618]. For a cyclostationary signal, its SCD function takes nonzero values at some nonzero cyclic frequencies. On the other hand, noise does not have any cyclostationarity at all; that is, its SCD function has zero values at all non-zero cyclic frequencies. Hence, we can distinguish signal from noise by analyzing the SCD function. Furthermore, it is possible to distinguish the signal type because different signals may have different non-zero cyclic frequencies.

In the following, we list cyclic frequencies for some signals of practical interest [17, 18].
  1. (1)

    Analog TV signal: it has cyclic frequencies at multiples of the TV-signal horizontal line-scan rate (15.75?KHz in USA, 15.625 KHz in Europe).

     
  2. (2)

    AM signal: https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq152_HTML.gif . It has cyclic frequencies at https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq153_HTML.gif .

     
  3. (3)

    PM and FM signal: https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq155_HTML.gif . It usually has cyclic frequencies at https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq156_HTML.gif . The characteristics of the SCD function at cyclic frequency https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq157_HTML.gif depend on https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq158_HTML.gif .

     
  4. (4)
    Digital-modulated signals are as follows
    1. (a)

      Amplitude-Shift Keying: https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq160_HTML.gif . It has cyclic frequencies at https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq161_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq162_HTML.gif

       
    2. (b)

      Phase-Shift Keying: https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq163_HTML.gif . For BPSK, it has cyclic frequencies at https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq164_HTML.gif , and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq165_HTML.gif For QPSK, it has cycle frequencies at https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq166_HTML.gif .

       
     

When source signal https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq167_HTML.gif passes through a wireless channel, the received signal is impaired by the unknown propagation channel. In general, the received signal can be written as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ34_HTML.gif
(34)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq168_HTML.gif denotes the convolution, and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq169_HTML.gif denotes the channel response. It can be shown that the SCD function of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq170_HTML.gif is

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ35_HTML.gif
(35)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq171_HTML.gif denotes the conjugate, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq172_HTML.gif denotes the cyclic frequency for https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq173_HTML.gif , https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq174_HTML.gif is the Fourier transform of the channel https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq175_HTML.gif , and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq176_HTML.gif is the SCD function of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq177_HTML.gif . Thus, the unknown channel could have major impacts on the strength of SCD at certain cyclic frequencies.

Although cyclostationary detection has certain advantages (e.g., robustness to uncertainty in noise power and propagation channel), it also has some disadvantages: https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq178_HTML.gif it needs a very high sampling rate; https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq179_HTML.gif the computation of SCD function requires large number of samples and thus high computational complexity; https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq180_HTML.gif the strength of SCD could be affected by the unknown channel; https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq181_HTML.gif the sampling time error and frequency offset could affect the cyclic frequencies.

8. Cooperative Sensing

When there are multiple users/receivers distributed in different locations, it is possible for them to cooperate to achieve higher sensing reliability, thus resulting in various cooperative sensing schemes [3444, 53, 62]. Generally speaking, if each user sends its observed data or processed data to a specific user, which jointly processes the collected data and makes a final decision, this cooperative sensing scheme is called data fusion. Alternatively, if multiple receivers process their observed data independently and send their decisions to a specific user, which then makes a final decision, it is called decision fusion.

8.1. Data Fusion

If the raw data from all receivers are sent to a central processor, the previously discussed methods for multi-antenna sensing can be directly applied. However, communication of raw data may be very expensive for practical applications. Hence, in many cases, users only send processed/compressed data to the central processor.

A simple cooperative sensing scheme based on the energy detection is the combined energy detection. For this scheme, each user computes its received source signal (including the noise) energy as https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq182_HTML.gif and sends it to the central processor, which sums the collected energy values using a linear combination (LC) to obtain the following test statistic:
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ36_HTML.gif
(36)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq183_HTML.gif is the combining coefficient, with https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq184_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq185_HTML.gif . If there is no information on the source signal power received by each user, the EGC can be used, that is, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq186_HTML.gif for all https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq187_HTML.gif . If the source signal power received by each user is known, the optimal combining coefficients can be found [38, 43]. For the low-SNR case, it can be shown [43] that the optimal combining coefficients are given by

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ37_HTML.gif
(37)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq188_HTML.gif is the received source signal (excluding the noise) power of user https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq189_HTML.gif .

A fusion scheme based on the CAV is given in [53], which has the capability to mitigate interference and noise uncertainty.

8.2. Decision Fusion

In decision fusion, each user sends its one-bit or multiple-bit decision to a central processor, which deploys a fusion rule to make the final decision. Specifically, if each user only sends one-bit decision ("1" for signal present and "0" for signal absent) and no other information is available at the central processor, some commonly adopted decision fusion rules are described as follows [42].
  1. (1)

    "Logical-OR (LO)" Rule: If one of the decisions is "1," the final decision is "1." Assuming that all decisions are independent, then the probability of detection and probability of false alarm of the final decision are https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq190_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq191_HTML.gif , respectively, where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq192_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq193_HTML.gif are the probability of detection and probability of false alarm for user https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq194_HTML.gif , respectively.

     
  2. (2)

    "Logical-AND (LA)" Rule: If and only if all decisions are "1," the final decision is "1." The probability of detection and probability of false alarm of the final decision are https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq195_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq196_HTML.gif , respectively.

     
  3. (3)
    " https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq197_HTML.gif out of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq198_HTML.gif " Rule: If and only if https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq199_HTML.gif decisions or more are "1"s, the final decision is "1." This includes "Logical-OR (LO)" ( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq200_HTML.gif ), "Logical-AND (LA)" ( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq201_HTML.gif ), and "Majority" ( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq202_HTML.gif ) as special cases [34]. The probability of detection and probability of false alarm of the final decision are
    https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ38_HTML.gif
    (38)

    respectively.

     

Alternatively, each user can send multiple-bit decision such that the central processor gets more information to make a more reliable decision. A fusion scheme based on multiple-bit decisions is shown in [41]. In general, there is a tradeoff between the number of decision bits and the fusion reliability. There are also other fusion rules that may require additional information [34, 63].

Although cooperative sensing can achieve better performance, there are some issues associated with it. First, reliable information exchanges among the cooperating users must be guaranteed. In an ad hoc network, this is by no means a simple task. Second, most data fusion methods in literature are based on the simple energy detection and flat-fading channel model, while more advanced data fusion algorithms such as cyclostationary detection, space-time combining, and eigenvalue-based detection, over more practical propagation channels need to be further investigated. Third, existing decision fusions have mostly assumed that decisions of different users are independent, which may not be true because all users actually receive signals from some common sources. At last, practical fusion algorithms should be robust to data errors due to channel impairment, interference, and noise.

9. Noise Power Uncertainty and Estimation

For many detection methods, the receiver noise power is assumed to be known a priori, in order to form the test statistic and/or set the test threshold. However, the noise power level may change over time, thus yielding the so-called noise uncertainty problem. There are two types of noise uncertainty: receiver device noise uncertainty and environment noise uncertainty. The receiver device noise uncertainty comes from [911]: (a) nonlinearity of receiver components and (b) time-varying thermal noise in these components. The environment noise uncertainty is caused by transmissions of other users, either unintentionally or intentionally. Because of the noise uncertainty, in practice, it is very difficult to obtain the accurate noise power.

Let the estimated noise power be https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq203_HTML.gif , where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq204_HTML.gif is called the noise uncertainty factor. The upper bound on https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq205_HTML.gif (in dB scale) is then defined as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ39_HTML.gif
(39)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq206_HTML.gif is called the noise uncertainty bound. It is usually assumed that https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq207_HTML.gif in dB scale, that is, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq208_HTML.gif , is uniformly distributed in the interval https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq209_HTML.gif [10]. In practice, the noise uncertainty bound of a receiving device is normally below 2 dB [10, 64], while the environment/interference noise uncertainty can be much larger [10]. When there is noise uncertainty, it is known that the energy detection is not effective [911, 64].

To resolve the noise uncertainty problem, we need to estimate the noise power in real time. For the multi-antenna case, if we know that the number of active primary signals, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq210_HTML.gif , is smaller than https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq211_HTML.gif , the minimum eigenvalue of the sample covariance matrix can be a reasonable estimate of the noise power. If we further assume to know the difference https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq212_HTML.gif , the average of the https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq213_HTML.gif smallest eigenvalues can be used as a better estimate of the noise power. Accordingly, instead of comparing the test statistics with an assumed noise power, we can compare them with the estimated noise power from the sample covariance matrix. For example, we can compare https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq214_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq215_HTML.gif with the minimum eigenvalue of the sample covariance matrix, resulting in the maximum to minimum eigenvalue (MME) detection and energy to minimum eigenvalue (EME) detection, respectively [21, 22]. These methods can also be used for the single-antenna case if signal samples are time-correlated [22].

Figures 1 and 2 show the Receiver Operating Characteristics (ROC) curves ( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq216_HTML.gif versus https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq217_HTML.gif ) at https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq218_HTML.gif  dB, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq219_HTML.gif , https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq220_HTML.gif , and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq221_HTML.gif . In Figure 1, the source signal is i.i.d and the flat-fading channel is assumed, while in Figure 2, the source signal is the wireless microphone signal [61, 65] and the multipath fading channel (with eight independent taps of equal power) is assumed. For Figure 2, in order to exploit the correlation of signal samples in both space and time, the received signal samples are stacked as in (27). In both figures, "ED-x dB" means the energy detection with x-dB noise uncertainty. Note that both BCED and ED use the true noise power to set the test threshold, while MME and EME only use the estimated noise power as the minimum eigenvalue of the sample covariance matrix. It is observed that for both cases of i.i.d source (Figure 1) and correlated source (Figure 2), BCED performs better than ED, and so does MME than EME. Comparing Figures 1 and 2, we see that BCED and MME work better for correlated source signals, while the reverse is true for ED and EME. It is also observed that the performance of ED degrades dramatically when there is noise power uncertainty.
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Fig1_HTML.jpg
Figure 1

ROC curve: i. i.d source signal.

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Fig2_HTML.jpg
Figure 2

ROC curve: wireless microphone source signal.

10. Detection Threshold and Test Statistic Distribution

To make a decision on whether signal is present, we need to set a threshold https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq222_HTML.gif for each proposed test statistic, such that certain https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq223_HTML.gif and/or https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq224_HTML.gif can be achieved. For a fixed sample size https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq225_HTML.gif , we cannot set the threshold to meet the targets for arbitrarily high https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq226_HTML.gif and low https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq227_HTML.gif at the same time, as they are conflicting to each other. Since we have little or no prior information on the signal (actually we even do not know whether there is a signal or not), it is difficult to set the threshold based on https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq228_HTML.gif . Hence, a common practice is to choose the threshold based on https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq229_HTML.gif under hypothesis https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq230_HTML.gif .

Without loss of generality, the test threshold can be decomposed into the following form: https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq231_HTML.gif , where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq232_HTML.gif is related to the sample size https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq233_HTML.gif and the target https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq234_HTML.gif , and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq235_HTML.gif is a statistic related to the noise distribution under https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq236_HTML.gif . For example, for the energy detection with known noise power, we have

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ40_HTML.gif
(40)

For the matched-filtering detection with known noise power, we have

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ41_HTML.gif
(41)

For the EME/MME detection with no knowledge on the noise power, we have

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ42_HTML.gif
(42)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq237_HTML.gif is the minimum eigenvalue of the sample covariance matrix. For the CAV detection, we can set

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ43_HTML.gif
(43)

In practice, the parameter https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq238_HTML.gif can be set either empirically based on the observations over a period of time when the signal is known to be absent, or analytically based on the distribution of the test statistic under https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq239_HTML.gif . In general, such distributions are difficult to find, while some known results are given as follows.

For energy detection defined in (8), it can be shown that for a sufficiently large values of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq240_HTML.gif , its test statistic can be well approximated by the Gaussian distribution, that is,

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ44_HTML.gif
(44)

Accordingly, for given https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq241_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq242_HTML.gif , the corresponding https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq243_HTML.gif can be found as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ45_HTML.gif
(45)

where

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ46_HTML.gif
(46)

For the matched-filtering detection defined in (9), for a sufficiently large https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq244_HTML.gif , we have

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ47_HTML.gif
(47)

Thereby, for given https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq245_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq246_HTML.gif , it can be shown that

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ48_HTML.gif
(48)

For the GLRT-based detection, it can be shown that the asymptotic (as https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq247_HTML.gif ) log-likelihood ratio is central chi-square distributed [13]. More precisely,

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ49_HTML.gif
(49)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq248_HTML.gif is the number of independent scalar unknowns under https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq249_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq250_HTML.gif . For instance, if https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq251_HTML.gif is known while https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq252_HTML.gif is not, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq253_HTML.gif will be equal to the number of independent real-valued scalar variables in https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq254_HTML.gif . However, there is no explicit expression for https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq255_HTML.gif in this case.

Random matrix theory (RMT) is useful for determining the test statistic distribution and the parameter https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq256_HTML.gif for the class of eigenvalue-based detection methods. In the following, we provide an example for the BCED detection method with known noise power, that is, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq257_HTML.gif . For this method, we actually compare the ratio of the maximum eigenvalue of the sample covariance matrix https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq258_HTML.gif to the noise power https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq259_HTML.gif with a threshold https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq260_HTML.gif . To set the value for https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq261_HTML.gif , we need to know the distribution of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq262_HTML.gif for any finite https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq263_HTML.gif . With a finite https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq264_HTML.gif , https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq265_HTML.gif may be very different from the actual covariance matrix https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq266_HTML.gif due to the noise. In fact, characterizing the eigenvalue distributions for https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq267_HTML.gif is a very complicated problem [6669], which also makes the choice of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq268_HTML.gif difficult in general.

When there is no signal, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq269_HTML.gif reduces to https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq270_HTML.gif , which is the sample covariance matrix of the noise only. It is known that https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq271_HTML.gif is a Wishart random matrix [66]. The study of the eigenvalue distributions for random matrices is a very hot research topic over recent years in mathematics, communications engineering, and physics. The joint PDF of the ordered eigenvalues of a Wishart random matrix has been known for many years [66]. However, since the expression of the joint PDF is very complicated, no simple closed-form expressions have been found for the marginal PDFs of the ordered eigenvalues, although some computable expressions have been found in [70]. Recently, Johnstone and Johansson have found the distribution of the largest eigenvalue [67, 68] of a Wishart random matrix as described in the following theorem.

Theorem 1.

Let https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq272_HTML.gif , https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq273_HTML.gif , and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq274_HTML.gif . Assume that https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq275_HTML.gif . Then, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq276_HTML.gif converges (with probability one) to the Tracy-Widom distribution of order 1 [71, 72].

The Tracy-Widom distribution provides the limiting law for the largest eigenvalue of certain random matrices [71, 72]. Let https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq277_HTML.gif be the cumulative distribution function (CDF) of the Tracy-Widom distribution of order 1. We have

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ50_HTML.gif
(50)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq278_HTML.gif is the solution of the nonlinear Painlevé II differential equation given by

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ51_HTML.gif
(51)

Accordingly, numerical solutions can be found for function https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq279_HTML.gif at different values of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq280_HTML.gif . Also, there have been tables for values of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq281_HTML.gif [67] and Matlab codes to compute them [73].

Based on the above results, the probability of false alarm for the BCED detection can be obtained as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ52_HTML.gif
(52)

which leads to

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ53_HTML.gif
(53)

or equivalently,

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ54_HTML.gif
(54)

From the definitions of https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq282_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq283_HTML.gif in Theorem 1, we finally obtain the value for https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq284_HTML.gif as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ55_HTML.gif
(55)

Note that https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq285_HTML.gif depends only on https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq286_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq287_HTML.gif . A similar approach like the above can be used for the case of MME detection, as shown in [21, 22].

Figure 3 shows the expected (theoretical) and actual (by simulation) probability of false alarm values based on the theoretical threshold in (55) for https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq288_HTML.gif , and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq289_HTML.gif . It is observed that the differences between these two sets of values are reasonably small, suggesting that the choice of the theoretical threshold is quite accurate.
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Fig3_HTML.jpg
Figure 3

Comparison of theoretical and actual https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq287_HTML.gif .

11. Robust Spectrum Sensing

In many detection applications, the knowledge of signal and/or noise is limited, incomplete, or imprecise. This is especially true in cognitive radio systems, where the primary users usually do not cooperate with the secondary users and as a result the wireless propagation channels between the primary and secondary users are hard to be predicted or estimated. Moreover, intentional or unintentional interference is very common in wireless communications such that the resulting noise distribution becomes unpredictable. Suppose that a detector is designed for specific signal and noise distributions. A pertinent question is then as follows: how sensitive is the performance of the detector to the errors in signal and/or noise distributions? In many situations, the designed detector based on the nominal assumptions may suffer a drastic degradation in performance even with small deviations from the assumptions. Consequently, the searching for robust detection methods has been of great interest in the field of signal processing and many others [7477]. A very useful paradigm to design robust detectors is the maxmin approach, which maximizes the worst case detection performance. Among others, two techniques are very useful for robust cognitive radio spectrum sensing: the robust hypothesis testing [75] and the robust matched filtering [76, 77]. In the following, we will give a brief overview on them, while for other robust detection techniques, the interested readers may refer to the excellent survey paper [78] and references therein.

11.1. Robust Hypothesis Testing

Let the PDF of a received signal sample be https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq291_HTML.gif at hypothesis https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq292_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq293_HTML.gif at hypothesis https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq294_HTML.gif . If we know these two functions, the LRT-based detection described in Section 2 is optimal. However, in practice, due to channel impairment, noise uncertainty, and interference, it is very hard, if possible, to obtain these two functions exactly. One possible situation is when we only know that https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq295_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq296_HTML.gif belong to certain classes. One such class is called the https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq297_HTML.gif -contamination class given by

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ56_HTML.gif
(56)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq298_HTML.gif ( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq299_HTML.gif ) is the nominal PDF under hypothesis https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq300_HTML.gif , https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq301_HTML.gif in https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq302_HTML.gif is the maximum degree of contamination, and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq303_HTML.gif is an arbitrary density function. Assume that we only know https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq304_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq305_HTML.gif (an upper bound for contamination), https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq306_HTML.gif . The problem is then to design a detection scheme to minimize the worst-case probability of error (e.g., probability of false alarm plus probability of mis-detection), that is, finding a detector https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq307_HTML.gif such that

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ57_HTML.gif
(57)
Hubber [75] proved that the optimal test statistic is a "censored" version of the LRT given by
https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ58_HTML.gif
(58)

where

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ59_HTML.gif
(59)

and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq308_HTML.gif , https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq309_HTML.gif are nonnegative numbers related to https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq310_HTML.gif , https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq311_HTML.gif , https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq312_HTML.gif , and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq313_HTML.gif [75, 78]. Note that if choosing https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq314_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq315_HTML.gif , the test is the conventional LRT with respect to nominal PDFs, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq316_HTML.gif and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq317_HTML.gif .

After this seminal work, there have been quite a few researches in this area [78]. For example, similar minmax solutions are found for some other uncertainty models [78].

11.2. Robust Matched Filtering

We turn the model (4) into a vector form as

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ60_HTML.gif
(60)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq318_HTML.gif is the signal vector and https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq319_HTML.gif is the noise vector. Suppose that https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq320_HTML.gif is known. In general, a matched-filtering detection is https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq321_HTML.gif . Let the covariance matrix of the noise be https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq322_HTML.gif . If https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq323_HTML.gif , it is known that choosing https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq324_HTML.gif is optimal. In general, it is easy to verify that the optimal https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq325_HTML.gif to maximize the SNR is

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ61_HTML.gif
(61)

In practice, the signal vector https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq326_HTML.gif may not be known exactly. For example, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq327_HTML.gif may be only known to be around https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq328_HTML.gif with some errors modeled by

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ62_HTML.gif
(62)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq329_HTML.gif is an upper bound on the Euclidean-norm of the error. In this case, we are interested in finding a proper value for https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq330_HTML.gif such that the worst-case SNR is maximized, that is,

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ63_HTML.gif
(63)

It was proved in [76, 77] that the optimal solution for the above maxmin problem is

https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_Equ64_HTML.gif
(64)

where https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq331_HTML.gif is a nonnegative number such that https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq332_HTML.gif .

It is noted that there are also researches on the robust matched filtering detection when the signal has other types of uncertainty [78]. Moreover, if the noise has uncertainties, that is, https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq333_HTML.gif is not known exactly, or both noise and signal have uncertainties, the optimal robust matched-filtering detector was also found for some specific uncertainty models in [78].

12. Practical Considerations and Future Developments

Although there have been quite a few methods proposed for spectrum sensing, their realization and performance in practical cognitive radio applications need to be tested [5052]. To build a practical sensing device, many factors should be considered. Some of them are discussed as follows.

( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq334_HTML.gif ) Narrowband noise. One or more narrowband filters may be used to extract the signal from a specific band. These filters can be analog or digital. Only if the filter is ideally designed and the signal is critically sampled (sampling rate is the same as the bandwidth of the filter), the discrete noise samples could be i.i.d. In a practical device, however, the noise samples are usually correlated. This will cause many sensing methods unworkable, because they usually assume that the noise samples are i.i.d. For some methods, a noise prewhitening process can be used to make the noise samples i.i.d. prior to the signal detection. For example, this method has been deployed in [22] to enable the eigenvalue-based detection methods. The similar method can be used for covariance-based detection methods, for example, the CAV.

( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq335_HTML.gif ) Spurious signal and interference. The received signal may contain not only the desired signal and white noise but also some spurious signal and interference. The spurious signal may be generated by Analog-to-Digital Converters (ADC) due to its nonlinearity [79] or other intentional/unintentional transmitters. If the sensing antenna is near some electronic devices, the spurious signal generated by the devices can be strong in the received signal. For some sensing methods, such unwanted signals will be detected as signals rather than noise. This will increase the probability of false alarm. There are methods to mitigate the spurious signal at the device level [79]. Alternatively, signal processing techniques can be used to eliminate the impact of spurious signal/interference [53]. It is very difficult, if possible, to estimate the interference waveform or distribution because of its variation with time and location. Depending on situations, the interference power could be lower or higher than the noise power. If the interference power is much higher than the noise power, it is possible to estimate the interference first and subtract it from the received signal. However, since we usually intend to detect signal at very low SNR, the error of the interference estimation could be large enough (say, larger than the primary signal) such that the detection with the residue signal after the interference subtraction is still unreliable. If the interference power is low, it is hard to estimate it anyway. Hence, in general we cannot rely on the interference estimation and subtraction, especially for very low-power signal detection.

( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq336_HTML.gif ) Fixed point realization. Many hardware realizations use fixed point rather than floating point computation. This will limit the accuracy of detection methods due to the signal truncation when it is saturated. A detection method should be robust to such unpredictable errors.

( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq337_HTML.gif ) Wideband sensing. A cognitive radio device may need to monitor a very large contiguous or noncontiguous frequency range to find the best available band(s) for transmission. The aggregate bandwidth could be as large as several GHz. Such wideband sensing requires ultra-wide band RF frontend and very fast signal processing devices. To sense a very large frequency range, typically a corresponding large sampling rate is required, which is very challenging for practical implementation. Fortunately, if a large part of the frequency range is vacant, that is, the signal is frequency-domain sparse, we can use the recently developed compressed sampling (also called compressed sensing) to reduce the sampling rate by a large margin [8082]. Although there have been studies in wideband sensing algorithms [26, 8387], more researches are needed especially when the center frequencies and bandwidths of the primary signals are unknown within the frequency range of interest.

( https://static-content.springer.com/image/art%3A10.1155%2F2010%2F381465/MediaObjects/13634_2009_Article_2768_IEq338_HTML.gif ) Complexity. This is of course one of the major factors affecting the implementation of a sensing method. Simple but effective methods are always preferable.

To detect a desired signal at very low SNR and in a harsh environment is by no means a simple task. In this paper, major attention is paid to the statistical detection methods. The major advantage of such methods is their little dependency on signal/channel knowledge as well as relative ease for realization. However, their disadvantage is also obvious: they are in general vulnerable to undesired interferences. How we can effectively combine the statistical detection with known signal features is not yet well understood. This might be a promising research direction. Furthermore, most exiting spectrum sensing methods are passive in the sense that they have neglected the interactions between the primary and secondary networks via their mutual interferences. If the reaction of the primary user (e.g., power control) upon receiving the secondary interference is exploited, some active spectrum sensing methods can be designed, which could significantly outperform the conventional passive sensing methods [88, 89]. At last, detecting the presence of signal is only the basic task of sensing. For a radio with high level of cognition, further information such as signal waveform and modulation schemes may be exploited. Therefore, signal identification turns to be an advanced task of sensing. If we could find an effective method for this advanced task, it in turn can help the basic sensing task.

13. Conclusion

In this paper, various spectrum sensing techniques have been reviewed. Special attention has been paid to blind sensing methods that do not need information of the source signals and the propagation channels. It has been shown that space-time joint signal processing not only improves the sensing performance but also solves the noise uncertainty problem to some extent. Theoretical analysis on test statistic distribution and threshold setting has also been investigated.

Authors’ Affiliations

(1)
Institute for Infocomm Research, A*STAR

References

  1. FCC : Spectrum policy task force report. Proceedings of the Federal Communications Commission (FCC '02), November 2002, Washington, DC, USA
  2. Islam MH, Koh CL, Oh SW, et al.: Spectrum survey in Singapore: occupancy measurements and analysis. Proceedings of the 3rd International Conference on Cognitive Radio Oriented Wireless Networks and Communications (CROWNCOM '08), May 2008, Singapor
  3. Mitola J, Maguire GQ: Cognitive radio: making software radios more personal. IEEE Personal Communications 1999, 6(4):13-18. 10.1109/98.788210View Article
  4. Haykin S: Cognitive radio: brain-empowered wireless communications. IEEE Transactions on Communications 2005, 23(2):201-220.
  5. Devroye N, Mitran P, Tarokh V: Achieveable rates in cognitive radio channels. IEEE Transactions on Information Theory 2006, 52(5):1813-1827.MathSciNetView ArticleMATH
  6. 802.22 Working Group : IEEE 802.22 D1: draft standard for wireless regional area networks. March 2008, http://​grouper.​ieee.​org/​groups/​802/​22/​
  7. Stevenson C, Chouinard G, Lei ZD, Hu WD, Shellhammer S, Caldwell W: IEEE 802.22: the first cognitive radio wireless regional area network standard. IEEE Communications Magazine 2009, 47(1):130-138.View Article
  8. Wireless Innovation Alliance 2008, http://​www.​wirelessinnovati​onalliance.​com/​
  9. Sonnenschein A, Fishman PM: Radiometric detection of spreadspectrum signals in noise of uncertainty power. IEEE Transactions on Aerospace and Electronic Systems 1992, 28(3):654-660. 10.1109/7.256287View Article
  10. Sahai A, Cabric D: Spectrum sensing: fundamental limits and practical challenges. IEEE International Symposium on New Frontiers in Dynamic Spectrum Access Networks (DySPAN '05), November 2005, Baltimore, Md, USA
  11. Tandra R, Sahai A: Fundamental limits on detection in low SNR under noise uncertainty. Proceedings of the International Conference on Wireless Networks, Communications and Mobile Computing (WirelessCom '05), June 2005, Maui, Hawaii, USA 1: 464-469.
  12. Tandra R, Sahai A: SNR walls for signal detection. IEEE Journal on Selected Topics in Signal Processing 2008, 2(1):4-17.View Article
  13. Kay SM: Fundamentals of Statistical Signal Processing: Detection Theory. Volume 2. Prentice Hall, Upper Saddle River, NJ, USA; 1998.
  14. Urkowitz H: Energy detection of unkown deterministic signals. Proceedings of the IEEE 1967, 55(4):523-531.View Article
  15. Chen HS, Gao W, Daut DG: Signature based spectrum sensing algorithms for IEEE 802.22 WRAN. Proceedings of the IEEE International Conference on Communications (ICC '07), June 2007, Glasgow, Scotland 6487-6492.
  16. Gardner WA: Exploitation of spectral redundancy in cyclostationary signals. IEEE Signal Processing Magazine 1991, 8(2):14-36. 10.1109/79.81007View Article
  17. Gardner WA: Spectral correlation of modulated signals—part I: analog modulation. IEEE Transactions on Communications 1987, 35(6):584-595. 10.1109/TCOM.1987.1096820View ArticleMATH
  18. Gardner WA, Brown WA III, Chen C-K: Spectral correlation of modulated signals—part II: digital modulation. IEEE Transactions on Communications 1987, 35(6):595-601. 10.1109/TCOM.1987.1096816View ArticleMATH
  19. Han N, Shon SH, Joo JO, Kim JM: Spectral correlation based signal detection method for spectrum sensing in IEEE 802.22 WRAN systems. Proceedings of the 8th International Conference on Advanced Communication Technology, Febraury 2006, Phoenix Park, South Korea
  20. Zeng YH, Liang Y-C: Eigenvalue based sensing algorithms. IEEE 802.22-06/0118r0, July 2006
  21. Zeng YH, Liang Y-C: Maximum-minimum eigenvalue detection for cognitive radio. Proceedings of the 18th International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC '07), September 2007, Athens, Greece
  22. Zeng YH, Liang Y-C: Eigenvalue-based spectrum sensing algorithms for cognitive radio. IEEE Transactions on Communications 2009, 57(6):1784-1793.View Article
  23. Bianchi P, Alfano JNG, Debbah M: Asymptotics of eigenbased collaborative sensing. Proceedings of the IEEE Information Theory Workshop (ITW '09), October 2009, Taormina, Sicily, Italy
  24. Maida M, Najim J, Bianchi P, Debbah M: Performance analysis of some eigen-based hypothesis tests for collaborative sensing. Proceedings of the IEEE Workshop on Statistical Signal Processing, September 2009, Cardiff, Wales, UK
  25. Penna F, Garello R, Spirito MA: Cooperative spectrum sensing based on the limiting eigenvalue ratio distribution in wishart matrices. IEEE Communications Letters 2009, 13(7):507-509.View Article
  26. Tian Z, Giannakis GB: A wavelet approach to wideband spectrum sensing for cognitive radios. Proceedings of the 1st International Conference on Cognitive Radio Oriented Wireless Networks and Communications (CROWNCOM '07), June 2007, Mykonos, Greece
  27. Zeng YH, Liang Y-C: Covariance based signal detections for cognitive radio. Proceedings of the 2nd IEEE International Symposium on New Frontiers in Dynamic Spectrum Access Networks (DySPAN '07), April 2007, Dublin, Ireland 202-207.View Article
  28. Zeng YH, Liang Y-C: Spectrum-sensing algorithms for cognitive radio based on statistical covariances. IEEE Transactions on Vehicular Technology 2009, 58(4):1804-1815.View Article
  29. Zeng YH, Liang Y-C, Zhang R: Blindly combined energy detection for spectrum sensing in cognitive radio. IEEE Signal Processing Letters 2008, 15: 649-652.View Article
  30. Pasupathy S, Venetsanopoulos AN: Optimum active array processing structure and space-time factorability. IEEE Transactions on Aerospace and Electronic Systems 1974, 10(6):770-778.View Article
  31. Dogandzic A, Nehorai A: Cramer-rao bounds for estimating range, velocity, and direction with an active array. IEEE Transactions on Signal Processing 2001, 49(6):1122-1137. 10.1109/78.923295View Article
  32. Fishler E, Haimovich A, Blum R, Chizhik D, Cimini L, Valenzuela R: MIMO radar: an idea whose time has come. Proceedings of the IEEE National Radar Conference, April 2004, Philadelphia, Pa, USA 71-78.
  33. Sheikhi A, Zamani A: Coherent detection for MIMO radars. Proceedings of the IEEE National Radar Conference, April 2007 302-307.
  34. Varshney PK: Distributed Detection and Data Fusion. Springer, New York, NY, USA; 1996.
  35. Unnikrishnan J, Veeravalli VV: Cooperative sensing for primary detection in cognitive radio. IEEE Journal on Selected Topics in Signal Processing 2008, 2(1):18-27.View Article
  36. Ganesan G, Li Y: Cooperative spectrum sensing in cognitive radio—part I: two user networks. IEEE Transactions on Wireless Communications 2007, 6: 2204-2213.View Article
  37. Ganesan G, Li Y: Cooperative spectrum sensing in cognitive radio—part II: multiuser networks. IEEE Transactions on Wireless Communications 2007, 6: 2214-2222.View Article
  38. Quan Z, Cui S, Sayed AH: Optimal linear cooperation for spectrum sensing in cognitive radio networks. IEEE Journal on Selected Topics in Signal Processing 2007, 2(1):28-40.View Article
  39. Mishra SM, Sahai A, Brodersen RW: Cooperative sensing among cognitive radios. Proceedings of the IEEE International Conference on Communications (ICC '06), June 2006, Istanbul, Turkey 4: 1658-1663.
  40. Sun C, Zhang W, Letaief KB: Cluster-based cooperative spectrum sensing in cognitive radio systems. Proceedings of the 18th International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC '07), September 2007, Athens, Greece 2511-2515.
  41. Ma J, Li Y: Soft combination and detection for cooperative spectrum sensing in cognitive radio networks. Proceedings of the IEEE Global Communications Conference (GlobeCom '07), November 2007, Washington, DC, USA
  42. Peh E, Liang Y-C: Optimization for cooperative sensing in cognitive radio networks. Proceedings of the IEEE Wireless Communications and Networking Conference (WCNC '07), March 2007, Hong Kong 27-32.
  43. Liang Y-C, Zeng YH, Peh E, Hoang AT: Sensing-throughput tradeoff for cognitive radio networks. IEEE Transactions on Wireless Communications 2008, 7(4):1326-1337.View Article
  44. Tandra R, Mishra SM, Sahai A: What is a spectrum hole and what does it take to recognize one. IEEE Proceedings 2009, 97(5):824-848.View Article
  45. Akyildiz IF, Lee WY, Vuran MC, Mohanty S: Next generation/dynamic spectrum access/cognitive radio wireless networks: a survey. Computer Networks Journal 2006, 50(13):2127-2159. 10.1016/j.comnet.2006.05.001View ArticleMATH
  46. Zhao Q, Sadler BM: A survey of dynamic spectrum access: signal processing, networking, and regulatory policy. IEEE Signal Processing Magazine 2007, 24: 79-89.View Article
  47. Ycek T, Arslan H: A survey of spectrum sensing algorithms for cognitive radio applications. IEEE Communications Surveys & Tutorials 2009, 11(1):116-160.View Article
  48. Sahai A, Mishra SM, Tandra R, Woyach KA: Cognitive radios for spectrum sharing. IEEE Signal Processing Magazine 2009, 26(1):140-145.View Article
  49. Ghasemi A, Sousa ES: Spectrum sensing in cognitive radio networks: requirements, challenges and design trade-offs. IEEE Communications Magazine 2008, 46(4):32-39.View Article
  50. Cabric D: Addressing the feasibility of cognitive radios. IEEE Signal Processing Magazine 2008, 25(6):85-93.View Article
  51. Oh SW, Naveen AAS, Zeng YH, et al.: White-space sensing device for detecting vacant channels in TV bands. Proceedings of the 3rd International Conference on Cognitive Radio Oriented Wireless Networks and Communications, (CrownCom '08), March 2008, Singapore
  52. Oh SW, Le TPC, Zhang WQ, Naveen SAA, Zeng YH, Kua KJM: TV white-space sensing prototype. Wireless Communications and Mobile Computing 2008, 9: 1543-1551.View Article
  53. Zeng YH, Liang Y-C, Peh E, Hoang AT: Cooperative covariance and eigenvalue based detections for robust sensing. Proceedings of the IEEE Global Communications Conference (GlobeCom '09), December 2009, Honolulu, Hawaii, USA
  54. Poor HV: An Introduction to Signal Detection and Estimation. Springer, Berlin, Germany; 1988.View ArticleMATH
  55. Van-Trees HL: Detection, Estimation and Modulation Theory. John Wiley & Sons, New York, NY, USA; 2001.View ArticleMATH
  56. Uchiyama H, Umebayashi K, Kamiya Y, et al.: Study on cooperative sensing in cognitive radio based ad-hoc network. Proceedings of the 18th International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC '07), September 2007, Athens, Greece
  57. Pandharipande A, Linnartz JPMG: Perfromcane analysis of primary user detection in multiple antenna cognitive radio. Proceedings of the IEEE International Conference on Communications (ICC '07), June 2007, Glasgow, Scotland
  58. Stoica P, Li J, Xie Y: On probing signal design for MIMO radar. IEEE Transactions on Signal Processing 2007, 55(8):4151-4161.MathSciNetView Article
  59. Giannakis GB, Hua Y, Stoica P, Tong L: Signal Processing Advances in Wireless & Mobile Communications. Prentice Hall PTR 2001., 1:
  60. Lim TJ, Zhang R, Liang YC, Zeng YH: GLRT-based spectrum sensing for cognitive radio. Proceedings of the IEEE Global Telecommunications Conference (GLOBECOM '08), December 2008, New Orleans, La, USA 4391-4395.
  61. Zeng YH, Liang Y-C: Simulations for wireless microphone detection by eigenvalue and covariance based methods. IEEE 802.22-07/0325r0, July 2007
  62. Peh E, Liang Y-C, Guan YL, Zeng YH: Optimization of cooperative sensing in cognitive radio networks: a sensing-throughput trade off view. IEEE Transactions on Vehicular Technology 2009, 58: 5294-5299.View Article
  63. Chair Z, Varshney PK: Optimal data fusion in multple sensor detection systems. IEEE Transactions on Aerospace and Electronic Systems 1986, 22(1):98-101.View Article
  64. Shellhammer S, Tandra R: Performance of the power detector with noise uncertainty. doc. IEEE 802.22-06/0134r0, July 2006
  65. Clanton C, Kenkel M, Tang Y: Wireless microphone signal simulation method. IEEE 802.22-07/0124r0, March 2007
  66. Tulino AM, Verdú S: Random Matrix Theory and Wireless Communications. Now Publishers, Hanover, Mass, USA; 2004.MATH
  67. Johnstone IM: On the distribution of the largest eigenvalue in principle components analysis. The Annals of Statistics 2001, 29(2):295-327. 10.1214/aos/1009210544MathSciNetView ArticleMATH
  68. Johansson K: Shape fluctuations and random matrices. Communications in Mathematical Physics 2000, 209(2):437-476. 10.1007/s002200050027MathSciNetView ArticleMATH
  69. Bai ZD: Methodologies in spectral analysis of large dimensional random matrices, a review. Statistica Sinica 1999, 9(3):611-677.MathSciNetMATH
  70. Zanella A, Chiani M, Win MZ: On the marginal distribution of the eigenvalues of wishart matrices. IEEE Transactions on Communications 2009, 57(4):1050-1060.View Article
  71. Tracy CA, Widom H: On orthogonal and symplectic matrix ensembles. Communications in Mathematical Physics 1996, 177(3):727-754. 10.1007/BF02099545MathSciNetView ArticleMATH
  72. Tracy CA, Widom H: The distribution of the largest eigenvalue in the Gaussian ensembles. In Calogero-Moser-Sutherland Models. Edited by: van Diejen J, Vinet L. Springer, New York, NY, USA; 2000:461-472.View Article
  73. Dieng MRMLab Version 0.02, 2006, http://​math.​arizona.​edu/​?​momar/​ RMLab Version 0.02, 2006,
  74. Huber PJ: Robust estimation of a location parameter. The Annals of Mathematical Statistics 1964, 35: 73-104. 10.1214/aoms/1177703732View ArticleMATHMathSciNet
  75. Huber PJ: A robust version of the probability ratio test. The Annals of Mathematical Statistics 1965, 36: 1753-1758. 10.1214/aoms/1177699803View ArticleMATHMathSciNet
  76. Zetterberg L-H: Signal detection under noise interference in a game situation. IRE Transactions on Information Theory 1962, 8: 47-57. 10.1109/TIT.1962.1057773View ArticleMATHMathSciNet
  77. Poor HV: Robust matched filters. IEEE Transactions on Information Theory 1983, 29(5):677-687. 10.1109/TIT.1983.1056734MathSciNetView ArticleMATH
  78. Kassam SA, Poor HV: Robust techniques for signal processing: a survey. Proceedings of the IEEE 1985, 73(3):433-482.View ArticleMATH
  79. Rabideau DJ, Howard LC: Mitigation of digital array nonlinearities. Proceedings of the IEEE National Radar Conference, May 2001, Atlanta, Ga, USA 175-180.
  80. Candès EJ, Romberg J, Tao T: Robust uncertainty principles: exact signal reconstruction from highly imcomplete frequency information. IEEE Transactions on Information Theory 2006, 52(2):489-509.View ArticleMATH
  81. Candès EJ, Wakin MB: An introduction to compressive sampling. IEEE Signal Processing Magazine 2008, 21(3):21-30.View Article
  82. Donoho DL: Compressed sensing. IEEE Transactions on Information Theory 2006, 52(4):1289-1306.MathSciNetView ArticleMATH
  83. Polo YL, Wang Y, Pandharipande A, Leus G: Compresive wideband spectrum sensing. Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP '09), April 2009, Taipei, Taiwan
  84. Zeng YH, Oh SW, Mo RH: Subcarrier sensing for distributed OFDMA in powerline communication. Proceedings of the IEEE International Conference on Communications (ICC '09), June 2009, Dresden, Germany
  85. Quan Z, Cui S, Poor HV, Sayed AH: Collaborative wideband sensing for cognitive radios. IEEE Signal Processing Magazine 2008, 25(6):60-73.View Article
  86. Quan Z, Cui S, Sayed AH, Poor HV: Optimal multiband joint detection for spectrum sensing in dynamic spectrum access networks. IEEE Transactions on Signal Processing 2009, 57(3):1128-1140.MathSciNetView Article
  87. Pei Y, Liang Y-C, Teh KC, Li KH: How much time is needed for wideband spectrum sensing? to appear in IEEE Transactions on Wireless Communications
  88. Zhang R, Liang Y-C: Exploiting hidden power-feedback loops for cognitive radio. Proceedings of the IEEE Symposium on New Frontiers in Dynamic Spectrum Access Networks (DySPAN '08), October 2008 730-734.
  89. Zhao G, Li YG, Yang C: Proactive detection of spectrum holes in cognitive radio. Proceedings of the IEEE International Conference on Communications (ICC '09), June 2009, Dresden, Germany

Copyright

© Yonghong Zeng et al. 2010

This article is published under license to BioMed Central Ltd. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.