Abstract-We present and discuss several novel applications of deep learning (DL) for the physical layer. By interpreting a communications system as an autoencoder, we develop a fundamental new way to think about communications system design as an end-to-end reconstruction task that seeks to jointly optimize transmitter and receiver components in a single process. We show how this idea can be extended to networks of multiple transmitters and receivers and present the concept of radio transformer networks (RTNs) as a means to incorporate expert domain knowledge in the machine learning (ML) model. Lastly, we demonstrate the application of convolutional neural networks (CNNs) on raw IQ samples for modulation classification which achieves competitive accuracy with respect to traditional schemes relying on expert features. The paper is concluded with a discussion of open challenges and areas for future investigation.

We study the adaptation of convolutional neural networks to the complex-valued temporal radio signal domain. We compare the efficacy of radio modulation classification using naively learned features against using expert feature based methods which are widely used today and e show significant performance improvements. We show that blind temporal learning on large and densely encoded time series using deep convolutional neural networks is viable and a strong candidate approach for this task especially at low signal to noise ratio.

We conduct an in depth study on the performance of deep learning based radio signal classification for radio communications signals. We consider a rigorous baseline method using higher order moments and strong boosted gradient tree classification and compare performance between the two approaches across a range of configurations and channel impairments. We consider the effects of carrier frequency offset, symbol rate, and multi-path fading in simulation and conduct over-the-air measurement of radio classification performance in the lab using software radios and compare performance and training strategies for both. Finally we conclude with a discussion of remaining problems, and design considerations for using such techniques.

We address the problem of learning efficient and adaptive ways to communicate binary information over an impaired channel. We treat the problem as reconstruction optimization through impairment layers in a channel autoencoder and introduce several new domain-specific regularizing layers to emulate common channel impairments. We also apply a radio transformer network based attention model on the input of the decoder to help recover canonical signal representations. We demonstrate some promising initial capacity results from this architecture and address several remaining challenges before such a system could become practical.

No abstract

IntroductionProviding optimal critical care in developing countries is limited by lack of recognition of critical illness and lack of essential resources. The Modified Early Warning Score (MEWS), based on physiological parameters, is validated in adult medical and surgical patients as a predictor of mortality. The objective of this study performed in Uganda was to determine the prevalence of critical illness on the wards as defined by the MEWS, to evaluate the MEWS as a predictor of death, and to describe additional risk factors for mortality.MethodsWe conducted a prospective observational study at Mulago National Referral Teaching Hospital in Uganda. We included medical and surgical ward patients over 18 years old, excluding patients discharged the day of enrolment, obstetrical patients, and patients who self-discharged prior to study completion. Over a 72-hour study period, we collected demographic and vital signs, and calculated MEWS; at 7-days we measured outcomes. Patients discharged prior to 7 days were assumed to be alive at study completion. Descriptive and inferential statistical analyses were performed.ResultsOf 452 patients, the median age was 40.5 (IQR 29–54) years, 53.3% were male, 24.3% were HIV positive, and 45.1% had medical diagnoses. MEWS ranged from 0 to 9, with higher scores representing hemodynamic instability. The median MEWS was 2 [IQR 1–3] and the median length of hospital stay was 9 days [IQR 4–24]. In-hospital mortality at 7-days was 5.5%; 41.4% of patients were discharged and 53.1% remained on the ward. Mortality was independently associated with medical admission (OR: 7.17; 95% CI: 2.064–24.930; p = 0.002) and the MEWS ≥ 5 (OR: 5.82; 95% CI: 2.420–13.987; p<0.0001) in the multivariable analysis.ConclusionThere is a significant burden of critical illness at Mulago Hospital, Uganda. Implementation of the MEWS could provide a useful triage tool to identify patients at greatest risk of death. Future research should include refinement of MEWS for low-resource settings, and development of appropriate interventions for patients identified to be at high risk of death based on early warning scores.

We introduce a novel physical layer scheme for single user Multiple-Input Multiple-Output (MIMO) communications based on unsupervised deep learning using an autoencoder. This method extends prior work on the joint optimization of physical layer representation and encoding and decoding processes as a single end-to-end task by expanding transmitter and receivers to the multi-antenna case. We introduce a widely used domain appropriate wireless channel impairment model (Rayleigh fading channel), into the autoencoder optimization problem in order to directly learn a system which optimizes for it. We considered both spatial diversity and spatial multiplexing techniques in our implementation. Our deep learning-based approach demonstrates significant potential for learning schemes which approach and exceed the performance of the methods which are widely used in existing wireless MIMO systems. We discuss how the proposed scheme can be easily adapted for open-loop and closed-loop operation in spatial diversity and multiplexing modes and extended use with only compact binary channel state information (CSI) as feedback.

This paper presents a novel method for synthesizing new physical layer modulation and coding schemes for communications systems using a learning-based approach which does not require an analytic model of the impairments in the channel. It extends prior work published on the channel autoencoder to consider the case where the channel response is not known or can not be easily modeled in a closed form analytic expression. By adopting an adversarial approach for channel response approximation and information encoding, we can jointly learn a good solution to both tasks over a wide range of channel environments. We describe the operation of the proposed adversarial system, share results for its training and validation over-the-air, and discuss implications and future work in the area.