Nowadays, with the proliferation of communication devices, Internet of Things (IoT) has become an emerging technology which facilitates massive devices to be enabled with connectivity by heterogeneous networks. However, it is usually a technical challenge for traditional networks to handle such a huge number of devices in an efficient manner. Recently, software defined network (SDN) technique with its agility and elasticity, has been incorporated into IoT to meet the potential scale and flexibility requirements, and forms a novel IoT architecture aka SDN-IoT. As the size of SDN-IoT increases, efficient routing protocols with low latency and high security are required, while the default routing protocols of SDN are still vulnerable to dynamic change of flow control rules especially when the network is under attack. To address the above issues, a Deep reinforcement learning based QoS-aware Secure routing Protocol (DQSP) is proposed in this paper. While guaranteeing the quality-of-service (QoS), our method can extract knowledge from history traffic demands by interacting with the underlying network environment, and dynamically optimize the routing policy. Extensive simulation experiments have been conducted with respect to several network performance metrics, demonstrating that our DQSP has good convergence and high effectiveness. Moreover, DQSP outperforms the traditional OSPF routing protocol, at least 10% relative performance gains in most cases.

As a novel architecture, software-defined networking (SDN) is viewed as the key technology of future networking. The core idea of SDN is to decouple the co

Subspace skyline, as an important variant of skyline, has been widely applied for multiple-criteria decisions, business planning. With the development of mobile internet, subspace skyline query in mobile distributed environments has recently attracted considerable attention. However, efficiently obtaining the meaningful subset of skyline points in any subspace remains a challenging task in the current mobile interne. For more and more mobile applications, subspace skyline query on mobile units is usually limited by big data and wireless bandwidth. To address this issue, in this paper, we propose a system model that can support subspace skyline query in mobile distributed environment. An efficient algorithm for processing the Subspace Skyline Query using MapReduce (SSQ) is also presented which can obtain the meaningful subset of points from the full set of skyline points in any subspace. The SSQ algorithm divides a subspace skyline query into two processing phases: the preprocess phase and the query phase. The preprocess phase includes the pruning process and constructing index process which is designed to reduce network delay and response time. Additionally, the query phase provides two filtering methods, SQM-filtering and epsilon-filtering, to filter the skyline points according to user preference and reduce network cost. Extensive experiments on real and synthetic data are conducted and the experimental results indicate that our algorithm is much efficient, meanwhile, the pruning strategy can further improve the efficiency of the algorithm. (C) 2015 Elsevier B.V. All rights reserved.

360° videos have been widely used with the development of virtual reality technology and triggered a demand to determine the most visually attractive objects in them, aka 360° video saliency prediction (VSP). While generative models, i.e., variational autoencoders or autoregressive models have proved their effectiveness in handling spatio-temporal data, utilizing them in 360° VSP is still challenging due to the problem of severe distortion and feature alignment inconsistency. In this study, we propose a novel spatio-temporal consistency generative network for 360° VSP. A dual-stream encoder-decoder architecture is adopted to process the forward and backward frame sequences of 360° videos simultaneously. Moreover, a deep autoregressive module termed as axial-attention based spherical ConvLSTM is designed in the encoder to memorize features with global-range spatial and temporal dependencies. Finally, motivated by the bias phenomenon in human viewing behavior, a temporal-convolutional Gaussian prior module is introduced to further improve the accuracy of the saliency prediction. Extensive experiments are conducted to evaluate our model and the state-of-the-art competitors, demonstrating that our model has achieved the best performance on the databases of PVS-HM and VR-Eyetracking.