Occurrence of periodical gapping in synthetic aperture radar (SAR) raw data poses a challenge in obtaining focused images. Traditional SAR imaging algorithms fail to process the periodically gapped raw data. A method is proposed to solve this problem. The presented method consists of phase compensation and recovering raw data in range Doppler domain via generalised orthogonal matching pursuit. Then, the recovered raw data is capable of being processed with the traditional SAR imaging algorithms. The Omega-K algorithm is selected to focus the reconstructed raw data. The presented method is validated on performing better in comparison with the spectral method periodically gapped amplitude and phase estimation by point target simulation and analysis. In simulation, the raw data loses 50% by the form of azimuth periodical gapping. The resolution of point targets can reach 0.25 m in both azimuth and range direction. In addition, real SAR raw data is utilised to verify the effectiveness of the proposed method.
Synthetic Aperture Radar (SAR) raw data missing occurs when radar is interrupted by various influences. In order to cope with this problem, a new method is proposed to focus the azimuth missing SAR raw data via segmented recovery in this paper. A reference function in time domain is designed to make the missing raw data sparser in two dimensional frequency domain. Afterwards, greedy algorithms are available to recover the missing data in two dimensional frequency domain. In addition, in order to avoid range frequency aliasing problem caused by reference function multiplication in time domain, the missing raw data is split into several parts in range direction and is recovered with a segmented recovery strategy. Then, the recovered raw data is available to be focused with traditional SAR imaging algorithms. The range migration algorithm is chosen to deal with the recovered raw data in this paper. Point target and area target simulations are carried out to validate the effectiveness of the proposed method on azimuth missing SAR raw data. Moreover, the proposed method is implemented on real SAR data in order to further provide convincing demonstration.
As one of the major search engines in the world, Baidu's Sponsored Search has long adopted the use of deep neural network (DNN) models for Ads click-through rate (CTR) predictions, as early as in 2013. The input futures used by Baidu's online advertising system (a.k.a. "Phoenix Nest") are extremely high-dimensional (e.g., hundreds or even thousands of billions of features) and also extremely sparse. The size of the CTR models used by Baidu's production system can well exceed 10TB. This imposes tremendous challenges for training, updating, and using such models in production. For Baidu's Ads system, it is obviously important to keep the model training process highly efficient so that engineers (and researchers) are able to quickly refine and test their new models or new features. Moreover, as billions of user ads click history entries are arriving every day, the models have to be retrained rapidly because CTR prediction is an extremely time-sensitive task. Baidu's current CTR models are trained on MPI (Message Passing Interface) clusters, which require high fault tolerance and synchronization that incur expensive communication and computation costs. And, of course, the maintenance costs for clusters are also substantial. This paper presents AIBox, a centralized system to train CTR models with tens-of-terabytes-scale parameters by employing solidstate drives (SSDs) and GPUs. Due to the memory limitation on GPUs, we carefully partition the CTR model into two parts: one is suitable for CPUs and another for GPUs. We further introduce a bi-level cache management system over SSDs to store the 10TB parameters while providing low-latency accesses. Extensive experiments on production data reveal the effectiveness of the new system. AIBox has comparable training performance with a large MPI cluster, while requiring only a small fraction of the cost for the cluster. CCS CONCEPTS • Information systems → Online advertising; • Hardware → Analysis and design of emerging devices and systems.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.