Automatic radio-frequency environment analysis

Abstract

The ability to automatically characterize all RF sources that have signijicant energy at a particular point in space has important applications in scientific, military, and industrial settings. Examples include automatic characterization of interference in radio astronomy, automatic signal detection and classijication for military surveillance, and inteverence characterization for communication-system test and evaluation. Such analyses are particularly diffcult when the unknown RF signals overlap in both time andfrequency or when the number of possible signal types is large.In this paper; we present a methd of automatically detecting, characterizing, and classijiing each of a number of RF sources that can spectrally and temporally overlap and that can be weak relative to the receiver noise. The method exploits the structure of higherorder statistics of manmade RF signals.

Robust automatic modulation classification and blind equalization: novel cognitive receivers

“…We first briefly describe nth order cumulant based AMC from [5,6]. We then propose a cost function that is related to the performance of the nth order cumulant based AMC.…”

“…Detailed tabulation can be found in [5,6]. From Table 1 it can be seen that the normalized cumulants values are unique for each modulation scheme and hence are used as a feature for classification.…”

“…Feature based AMC algorithms offer good performance and are relatively simple to implement. Feature based AMC algorithms that utilize nth order cumulant features are widely used because these features can distinguish a wide variety of modulation schemes and are easy to implement [4][5][6]. In a typical wireless communication environment, the transmitted signals are subjected to noise and multipath fading.…”

Spectrum sensing for cognitive radio: A signal-processing perspective on signal-statistics exploitation

“…Blind signal detection, classification, and parameter estimation can be performed by properly exploiting the statistics exhibited by communication signals [11], which are ignored by ED. Almost all communication signals are cyclostationary signals, which means they have one or more nth-order moment functions that are periodic or almost periodic in time for n ≥ 2 [13].…”

A Hiearchical framework for Signal Detection in Cognitive Radio Networks using SVM and Decision Tree Approaches

“…Practically, all types of communication signals have periodic properties where as noise is random in behavior. Cyclostationary feature based signal detection and classification helps to detect the signal type of a wide range of unknown signals (No prior knowledge of signal characteristics like the carrier frequency, phase or symbol rate) [6]- [8]. With the knowledge different cyclostationary features of different modulation schemes learning techniques can be employed to classify the same.…”