No abstract

Queue-aware energy-efficient scheduling and power allocation with feedback reduction in small-cell networks SCIENCE CHINA Information Sciences 61, 048301 (2018); A suboptimal joint bandwidth and power allocation for cooperative relay networks: a cooperative game theoretic approach SCIENCE CHINA Information Sciences 56, 072304 (2013); An energy-efficient geographic routing based on cooperative transmission in wireless sensor networks SCIENCE CHINA Information Sciences 56, 072302 (2013); Energy-efficient power allocation for non-regenerative OFDM relay links SCIENCE CHINA Information Sciences 56, 022306 (2013);. RESEARCH PAPER. SCIENCE CHINA Information Sciences

No abstract

No abstract

Coalfield subsurface fires can result in ecological disasters of global dimensions. These fires are difficult to control therefore can result in colossal wastage of resources (the coal itself but the resources devoted to suppression), a serious negative impact on the environment and acute health problems for large populations. However, if the heat can be effectively recycled and utilized, the combustion energy will be recovered but also heat extraction can promote suppression. Thus, leading not only to a positive energy impact but to a reduction polluting emissions and consequent health issues. This paper presents the results of a feasibility analysis of the overall recovery of underground thermal resources of a novel system of Waste Heat Recovery Units (WHRUS) that combines thermosyphon and thermoelectric technologies. Both thermal equivalent model and numerical assessment are presented. A series of realistic-scale field experiment conducted in the Xinjiang's fire zone for an extended period are discussed. Using a local geothermic assessment, the heat recovered from subsurface coal fire can be estimated as the summation of the convective and conductive components of the energy generated. The average heat generated for the fire district is estimated at approximately 495 W/m 2 and the average extraction efficiency at approximately 58%. The WHRUS shows and excellent heat transfer performance with an effective lower resistance of approximately 0.0049 W/°C and maximum thermal recovery rate greater than 90%. Finally, while the thermoelectric

D2D communication has been proposed as an important supplement to the existing centralized cellular networks which allows two physically adjacent cellular user equipments (UEs) to communicate directly. This paper concerns using D2D to improve wireless multicast services in cellular networks. Specially, we consider a D2D transmitter UE can act as a full-duplex (FD) relay to assist a cellular multicast from a base station (BS) to a group of two UEs. And a new scheme which allows an intra-cell D2D retransmission to underlay a cellular multicast is proposed. Under the constraint of the minimum signal-to-interference-and-noise ratio (SINR) required by each of the receiver UEs, the aim of the scheme is to select the best UE in a multicast group to perform the D2D retransmission with the serving BS. Thus, the aggregate transmit power consumed at the BS and at the selected UE can be minimized. The numerical results show that the proposed scheme outperforms traditional cellular multicast scheme as it consumes less transmit power to achieve the same SINR target at the receiver UEs.

Extreme learning machine (ELM) has been proposed for solving fast supervised learning problems by applying random computational nodes in the hidden layer. Similar to support vector machine, ELM cannot handle high-dimensional data effectively. Its generalization performance tends to become bad when it deals with highdimensional data. In order to exploit high-dimensional data effectively, a two-stage extreme learning machine model is established. In the first stage, we incorporate ELM into the spectral regression algorithm to implement dimensionality reduction of high-dimensional data and compute the output weights. In the second stage, the decision function of standard ELM model is computed based on the lowdimensional data and the obtained output weights. This is due to the fact that two stages are all based on ELM. Thus, output weights in the second stage can be approximately replaced by those in the first stage. Consequently, the proposed method can be applicable to high-dimensional data at a fast learning speed. Experimental results show that the proposed two-stage ELM scheme tends to have better scalability and achieves outstanding generalization performance at a faster learning speed than ELM.

No abstract