In this paper, we investigate the physical-layer security of cooperative communications relying on multiple two-way relays using decode-and-forward (DF) protocol in presence an eavesdropper, where eavesdropper appears to tap transmissions both source and relay. The design tradeoff be resolved is that throughput improved by invoking relaying, but secrecy wireless may degraded, since overhear signals transmitted relay nodes. We conceive artificial noise aided opportunistic selection (ANaTWORS)...

By reason of factors such as terrains, weather conditions, sensor imaging methods and cultural economic development, there is a large shift between the remote sensing imagery collected from different geographic locations sensors, which makes state-of-the-art semantic segmentation models trained on source domain (a image set gathered specific sensors) difficult to generalize target (another other sensors). Currently, unsupervised adaptation using adversarial training whose purpose align...

We study the physical-layer security of a cognitive radio system in face multiple eavesdroppers (EDs), which is composed secondary base station (SBS), users (SUs) as well pair primary transmitter (PT) and receiver (PR), where SUs first harvest energy from their received frequency signals transmitted by PT then communicate with SBS relying on opportunistic scheduling. consider two specific user scheduling schemes, namely, channel-aware (CaUS) energy-aware (EaUS). In CaUS scheme, an SU having...

Semantic segmentation of small-scale objects in very high resolution (VHR) remote sensing images plays an important role some special tasks, such as change detection and mapping land cover. However, due to small size, are more likely be completely obscured by shadows than large-scale objects, which make it difficult for the traditional convolutional neural network (CNN) distinguish from shadows. Furthermore, even if distinguished, their boundaries still refine. To solve above problems, a...

Fingerprint localization(FL) is one of the most efficient positioning scheme which exploits characteristics received signal or channel information to estimate physical position. Although there are many available techniques, them used in indoor positioning. In this paper, we discuss a possible method locate mobile device massive multiple-in-multiple-out(MIMO) systems represents leading 5G technology candidate. offline phase, fingerprint matrix based on angle-delay power extracted and...

Remarkable improvements have been seen in the semantic segmentation of remote-sensing images. As an effective structure to aggregate shallow information and deep information, encoder–decoder has widely used many state-of-the-art models, but it possesses two drawbacks that not fully addressed. On one hand, fuses features obtained from layers directly; despite harvesting some detailed also brings noisy owing poor discriminant ability layers. other existing merely high-level generated by last...

Global context information is vital in visual understanding problems, especially pixel-level semantic segmentation. The mainstream methods adopt the self-attention mechanism to model global information. However, pixels belonging different classes usually have weak feature correlation. Modeling correlation matrix indiscriminately extremely redundant mechanism. In order solve above problem, we propose a hierarchical network differentially homogeneous with strong correlations and heterogeneous...

Between the two broad categories of packet scheduling algorithms, round robin and time stamp based schemes, former is usually not considered a proper candidate for providing QoS guarantees. In this paper, we challenge such conventional wisdom by comparing them with new more balanced perspective. Taking scheduler provisioning condition link as whole, it shown extensive simulations that RR algorithms little bandwidth overprovisioning can beat TS ones. Considering prospect overprovisioning,...