Among the proliferation of device-to-device (D2D) based caching schemes, the most popular content (MPC) caching and the largest content diversity (LCD) caching have recently gained research attention. By jointly exploiting the merits of both MPC, LCD and grouped caching strategies, in this paper, we propose a new multi-popularity multi-threshold grouped caching scheme for D2D communication (MMGC-D2D). Specifically, we first devise an optimal device grouping scheme aiming to improve the regional user fairness of grouped caching. Then, to maximize the gain of cache space and to minimize the outage probability of cached packets, we discuss in details a novel MPC-LCD threshold-based grouped caching placement optimization problem and its analytical property, while taking into account the packet request diversity (PRD). Simulations results show that the proposed scheme exhibits notable gains over existing methods.

Device-to-device (D2D) based packet caching technologies recently attract increasing attention, thanks to their great potentials to facilitate network traffic offloading. Despite the many popular issues arising in the D2D caching area, one of the new perspectives, namely the competition due to the packet request diversity originating from various D2D user equipment (UE) groups is not sufficiently investigated. In this work, we analyze several key aspects of the competition for packet allocation among diverse packet requests. Firstly, we study the impact from diverse group proportions on the system throughput and the packet allocation fairness. Particularly, the novel group separation index (GSI) is introduced, which helps to reflect the packet allocation fairness. We derive and analyze both the upper and lower bounds of GSI. Secondly, we investigate how the concentration levels of diverse packet requests may affect the system performance, such as the impact from the caching size limit on packet allocation. Thirdly, we derive the average energy consumption metric using the binomial point process based network model, which facilitates a comprehensive evaluation of the competition among UE groups. Finally, simulations validate our proposed analysis method, which may provide important design hints for improving existing D2D caching schemes.

Ultrasonography is one of the main imaging methods for diagnosing thyroid nodules. Automatic differentiation between benign and malignant nodules in ultrasound images can greatly assist inexperienced clinicians in their diagnosis. The key of problem is the effective utilization of the features of ultrasound images. In this study, we propose a method that is based on the combination of conventional ultrasound and ultrasound elasticity images based on a convolutional neural network and introduces richer feature information for the classification of benign and malignant thyroid nodules. First, the conventional network model performs pretraining on ImageNet and transfers the feature parameters to the ultrasound image domain by transfer learning so that depth features may be extracted and small samples may be processed. Then, we combine the depth features of conventional ultrasound and ultrasound elasticity images to form a hybrid feature space. Finally, the classification is completed on the hybrid feature space, and an end-to-end CNN model is implemented. The experimental results demonstrate that the accuracy of the proposed method is 0.9470, which is better than that of other single data-source methods under the same conditions.

Applying sociality-aware (SA) device-to-device (D2D) user equipment (UE)-to-network (UE-NW) relay can help to improve UEs' accessibility to the network, thus increasing the network coverage and capacity. In this context, one key issue is the design of the relay helper UE (HUE) selection procedure. However, few existing studies jointly consider the diverse impacts from various design criteria on the performance of HUE selection. In this paper, the sociality-aware D2D UE-NW relay selection (RS) issue is formulated as a constrained multi-objective binary integer linear programming (MOBILP) problem, which can be solved by the proposed SA-D2D-RS method. Specifically, the new solution jointly exploits the fuzzy analytic hierarchy process (FAHP) and the entropy weight generation method that provides both subjective and objective preferences in decision making. Simulation results validate the superiority of the proposed scheme through realistic social data traces, and demonstrate its high access ratio as well as system throughput.

Device-to-device (D2D) multicast retransmissions in cellular networks helps to improve spectral efficiency and to offload traffic. Existing methods have revealed the relationship between the number of relays and resource utilisation efficiency on retransmissions in terms of relay selection and clustering. However, few studies consider the impact from the lost packets on resource utilisation, which is however important for improving the overall efficiency of D2D systems. In this paper, we first formulate a general resource efficiency optimisation problem in D2D multicast retransmission scenario, where the lost-packet diversity (LPD) of remote user equipments (UE) and link quality of D2D links are jointly considered. Then, we propose an efficient near-optimal algorithm to solve the new optimisation problem. Simulation results validate the benefits of the proposed algorithm which provides a near-optimal performance at low complexity.

Mining subsidence brings enormous damages in mining area. Facing this problem, the paper proposes a solution based on integrated GIS. It takes the coal mining, terrain motion, subsidence predication, surface destruction and disaster evaluation as a whole integrated system. The solution provides an assistant decision support for mining subsidence's prevention and curing by means of classic probability integral model and GIS components model. Based on the research and analysis, a mining subsidence predication & assistant decision support system (abbreviated as MSPADSS) is developed using oriented object programming and component GIS development technology. MSPADSS is composed of five modules, i.e., data collection, mining subsidence predication, GIS graphic and attribute information management, surface subsidence damage evaluation and GIS spatial analysis & thematic mapping. All these are integrated in a whole system. Several recognitions and conclusions can be acquired through the research as follows: Firstly, prevention and curing of mining subsidence should consider many relative factors, which cover features from the bottom of the mine to the ground. Secondly, the subsidence prediction and damage evolution can be calculated and visualized synchronously by means of ComGIS developing. Thirdly, application of MSPADSS in the east of China shows that this integrated GIS solution is feasible and adoptable.