Automatic modulation classification (AMC), which plays critical roles in both civilian and military applications, is investigated in this paper through a deep learning approach. Conventional AMCs can be categorized into maximum likelihood (ML) based (ML-AMC) and feature-based AMC. However, the practical deployment of ML-AMCs is difficult due to its high computational complexity, and the manually extracted features require expert knowledge. Therefore, an end-to-end convolution neural network (CNN) based AMC (CNN-AMC) is proposed, which automatically extracts features from the long symbol-rate observation sequence along with the estimated signal-to-noise ratio (SNR). With CNN-AMC, a unit classifier is adopted to accommodate the varying input dimensions. The direct training of CNN-AMC is challenging with the complicated model and complex tasks, so a novel two-step training is proposed, and the transfer learning is also introduced to improve the efficiency of retraining. Different digital modulation schemes have been considered in distinct scenarios, and the simulation results show that the CNN-AMC can outperform the feature-based method, and obtain a closer approximation to the optimal ML-AMC. Besides, CNN-AMCs have the certain robustness to estimation error on carrier phase offset and SNR. With parallel computation, the deep-learning-based approach is about $ 40$ to $ 1700$ times faster than the ML-AMC regarding inference speed.

Automatic Modulation Classification: A Deep Learning Enabled Approach

Adaptive modulation and automatic modulation classification are highly demanded in software-defined radio (SDR) for both commercial and military applications. Various design options of automatic classifiers have attracted researchers in developing 3G and 4G wireless communication systems. There is an urgent need to investigate the different methods of coherent and noncoherent modulation estimations, discuss the challenges in cooperative and noncooperative communication environment, and understand the distinct requirements in real-time modulation classifications. This survey paper focuses on the automatic modulation classification methods based on likelihood functions, studies various classification solutions derived from likelihood ratio test, and discusses the detailed characteristics associated with all major algorithms.

Likelihood-Ratio Approaches to Automatic Modulation Classification

The Internet of Things (IoT) is expected to require more effective and efficient wireless communications than ever before. For this reason, techniques such as spectrum sharing, dynamic spectrum access, extraction of signal intelligence and optimized routing will soon become essential components of the IoT wireless communication paradigm. In this vision, IoT devices must be able to not only learn to autonomously extract spectrum knowledge on-the-fly from the network but also leverage such knowledge to dynamically change appropriate wireless parameters ( e.g. , frequency band, symbol modulation, coding rate, route selection, etc.) to reach the network’s optimal operating point. Given that the majority of the IoT will be composed of tiny, mobile, and energy-constrained devices, traditional techniques based on a priori network optimization may not be suitable, since (i) an accurate model of the environment may not be readily available in practical scenarios; (ii) the computational requirements of traditional optimization techniques may prove unbearable for IoT devices. To address the above challenges, much research has been devoted to exploring the use of machine learning to address problems in the IoT wireless communications domain. The reason behind machine learning’s popularity is that it provides a general framework to solve very complex problems where a model of the phenomenon being learned is too complex to derive or too dynamic to be summarized in mathematical terms. This work provides a comprehensive survey of the state of the art in the application of machine learning techniques to address key problems in IoT wireless communications with an emphasis on its ad hoc networking aspect. First, we present extensive background notions of machine learning techniques. Then, by adopting a bottom-up approach, we examine existing work on machine learning for the IoT at the physical, data-link and network layer of the protocol stack. Thereafter, we discuss directions taken by the community towards hardware implementation to ensure the feasibility of these techniques. Additionally, before concluding, we also provide a brief discussion of the application of machine learning in IoT beyond wireless communication. Finally, each of these discussions is accompanied by a detailed analysis of the related open problems and challenges.

Machine learning for wireless communications in the Internet of Things: A comprehensive survey

Signal identification, which initially found applications in electronic warfare and spectrum monitoring and surveillance, has been recently considered for commercial communications in the context of software defined and cognitive radios. In this article, I present a snapshot of the status of signal identification algorithms, starting from a general description of maximum likelihood (ML) and feature based (FB) approaches to a more detailed discussion of a practical methodology using cyclostationarity-based features. I discuss the cyclostationarity-based features of various signals and the criteria of decision for their identification, while considering classical problems of identifying single carrier linearly digitally (SCLD) modulated signals, as well as new challenges posed by the identification of orthogonal frequency division multiplexing (OFDM), SC frequency domain equalization (SC-FDE), and multiple-transmit antenna signals. I conclude the article with remarks on practical solutions to signal identification and open research issues.

Signal identification for emerging intelligent radios: classical problems and new challenges

This paper presents an overview of feature-based (FB) methods developed for Automatic classification of digital modulations. Only the most well-known features and classifiers are considered, categorized, and defined. The features include instantaneous time domain (ITD) parameters, Fourier transform (FT), wavelet transform (WT), higher order moments (HOM) to name a few. The classifiers are artificial neural networks (ANN), support vector machines (SVMs), and decision tree (DT). We also highlight the advantages and disadvantages of each technique in classifying a certain modulation scheme. The objective of this work is to assist newcomers to the field to choose suitable algorithms for intended applications. Furthermore, this work is expected to help in determining the limitations associated with the available FB automatic modulation classification (AMC) methods.

An overview of feature-based methods for digital modulation classification

Automatic modulation recognition (AMR)-based software-defined radio (SDR) is a research challenge in developing third-generation (3G) and fourth-generation (4G) wireless communications with adaptive modulation capability. However, the existing AMR technology does not satisfy the seamless demodulation requirement of the SDR. A novel design of the AMR method with reduced computational complexity and fast processing speed is needed. This paper describes a discrete likelihood-ratio test (DLRT)-based rapid-estimation approach to identifying the modulation schemes blindly for uninterrupted data demodulation in real time. The statistical performance of the fast AMR associated with its implementation using the SDR is presented.

Software-Defined Radio Equipped With Rapid Modulation Recognition

This paper describes a likelihood test based modulation classification method for identifying the modulation scheme of a software-defined radio (SDR) in real-time without pilot symbols between transmitters and receivers. Unlike the prior art, the paper converts an unknown signal symbol to an address of the look-up table (LUT), loads the pre-calculated values of the test functions for the likelihood ratio test, and produces the estimated modulation scheme in real-time. The statistical performance of the LUT based classifier is studied. Simulation results are presented to confirm the theoretical analysis.

Real-time Modulation Classification Based On Maximum Likelihood

Automatic modulation classification (AMC) has been intensively studied to enhance the successful classification rate, particularly for overcoming the physical limit that deals with weak signals received in a noncooperative communication environment. A wireless sensor network (WSN) has multiple geometrically distributed sensors to work cooperatively. The distributed signal sensing and classification performed by collaborated sensors is proven to be beneficial to increasing the modulation classification reliability. In this paper, we apply the likelihood ratio-based distributed detection fusion technique to address the issues of general binary modulation classifications. The data fusion algorithm performed in the primary node is presented. Its numerical performance with simulation results is demonstrated.

Distributed Automatic Modulation Classification With Multiple Sensors

Automatic modulation classification with a single receiver has been intensively studied for two decades. Enhancing the successful classification probability is a bottleneck in this research field especially with weak signals in a non-cooperative communication environment. A sensor network with distributed classification techniques is expected to achieve technology breakthrough in providing spatial diversity and increasing the classification reliability. In this paper, we developed a distributed likelihood function-based classification method and extend the automatic modulation classification to sensor or radio networks. The classification methods performed in the sensors and primary node associated with theoretical discussion and numerical results are presented.

Jefferson L. Xu

Papers

Likelihood-Ratio Approaches to Automatic Modulation Classification

Software-Defined Radio Equipped With Rapid Modulation Recognition

Real-time Modulation Classification Based On Maximum Likelihood

Distributed Automatic Modulation Classification With Multiple Sensors

Likelihood function-based modulation classification in bandwidth-constrained sensor networks