With the rapid increase in demand for mobile data, mobile network operators are trying to expand wireless network capacity by deploying wireless local area network (LAN) hotspots on to which they can offload their mobile traffic. However, these network-centric methods usually do not fulfill the interests of mobile users (MUs). Taking into consideration many issues such as different applications' deadlines, monetary cost and energy consumption, how the MU decides whether to offload their traffic to a complementary wireless LAN is an important issue. Previous studies assume the MU's mobility pattern is known in advance, which is not always true. In this paper, we study the MU's policy to minimize his monetary cost and energy consumption without known MU mobility pattern. We propose to use a kind of reinforcement learning technique called deep Q-network (DQN) for MU to learn the optimal offloading policy from past experiences. In the proposed DQN based offloading algorithm, MU's mobility pattern is no longer needed. Furthermore, MU's state of remaining data is directly fed into the convolution neural network in DQN without discretization. Therefore, not only does the discretization error present in previous work disappear, but also it makes the proposed algorithm has the ability to generalize the past experiences, which is especially effective when the number of states is large. Extensive simulations are conducted to validate our proposed offloading algorithms.

This paper presents a high power efficient pulse VCO with tail-filter for the chip-scale atomic clock (CSAC) application. The stringent power and clock stability specifications of next-generation CSAC demand a VCO with ultra-low power consumption and low phase noise. The proposed VCO architecture aims for the high power efficiency, while further reducing the phase noise using tail filtering technique. The VCO has been implemented in a standard 45nm SOI technology for validation. At an oscillation frequency of 5.0GHz, the proposed VCO achieves a phase noise of -120dBc/Hz at 1MHz offset, while consuming 1.35mW. This translates into an FoM of -191dBc/Hz.

Tracking-by-detection methods consider tracking task as a continuous detection problem applied over video frames. Modern tracking-by-detection trackers have online learning ability; the update stage is essential because it determines how to modify the classifier inherent in a tracker. However, most trackers search for the target within a fixed region centered at the previous object position; thus, they lack spatiotemporal consistency. This becomes a problem when the tracker detects an incorrect object during short-term occlusion. In addition, the scale of the bounding box that contains the target object is usually assumed not to change. This assumption is unrealistic for long-term tracking, where the scale of the target varies as the distance between the target and the camera changes. The accumulation of errors resulting from these shortcomings results in the drift problem, i.e. drifting away from the target object. To resolve this problem, we present a drift-free, online learning-based tracking-by-detection method using a single static camera. We improve the latent structured support vector machine (SVM) tracker by designing a more robust tracker update step by incorporating two Kalman filter modules: the first is used to predict an adaptive search region in consideration of the object motion; the second is used to adjust the scale of the bounding box by accounting for the background model. We propose a hierarchical search strategy that combines Bhattacharyya coefficient similarity analysis and Kalman predictors. This strategy facilitates overcoming occlusion and increases tracking efficiency. We evaluate this work using publicly available videos thoroughly. Experimental results show that the proposed method outperforms the state-of-the-art trackers.

The unconditional stable finite-difference time-domain (US-FDTD) method based on Laguerre polynomial expansion and Galerkin temporal testing is used to model thin-film bulk acoustic wave resonators (TFBAR). Numerical results show the efficiency of the US-FDTD algorithm.

Sequences that attain the smallest possible absolute sidelobes (SPASs) of periodic autocorrelation function (PACF) play fairly important roles in synchronization of communication systems, Large scale integrated circuit testing, and so on. This letter presents an approach to construct 16-QAM sequences of even periods, based on the known quaternary sequences. A relationship between the PACFs of 16-QAM and quaternary sequences is established, by which when quaternary sequences that attain the SPASs of PACF are employed, the proposed 16-QAM sequences have good PACF.

This paper presents a fully integrated yet compact receiver front-end for Sub-GHz applications such as Internet-of-Things (IoT). The low noise amplifier (LNA) matching network leverages an inductance boosting technique. A relatively small on-chip inductor with a compact area achieves impedance matching in such a low frequency. Moreover, a passive-mixer-first mode bypasses the LNA to extend the receiver dynamic-range. The passive mixer provides matching to the 50Ω antenna interface to eliminate the need for additional passive components. Therefore, the receiver can be fully-integrated without any off-chip matching components. The flipped-voltage-follower (FVF) cell is adopted in the low pass filter (LPF) and the variable gain amplifier (VGA) for its high linearity and low power consumption. Fabricated in 65nm LP CMOS process, the proposed receiver front-end occupies 0.37mm2 core area, with a tolerable input power ranging from -91.5dBm to -1dBm for 500kbps GMSK signal at 924MHz frequency. The power consumption is 1mW power under a 1.2V supply.

Parallel TCP is one possible approach to increasing throughput of data transfer in Long Fat Networks (LFNs). Using parallel TCP is something of black art. As high-speed transport-layer protocols appear, e.g. HSTCP, it is necessary to reinvestigate the performance of parallel TCP, because a choice has to be make among them for the system. In this paper, the performance of parallel TCP is evaluated by mathematical analysis based on a simple dumbbell topology. Packet drop rate and aggregate goodput are used as two metrics to characterize the performance of parallel TCP. Two cases, namely synchronization and non-synchronization, are analyzed in detail when DropTail is deployed on routers. The synchronization case is common in using parallel TCP, but the goodput deteriorates seriously. The non-synchronization case may benefit parallel TCP, but extra mechanisms are required, and it is not easy to implement in the real world. The problem also remains even if Random Early Detection (RED) queue management is employed on routers. The analysis results show the difficulty in using parallel TCP in practice.

In this paper, we propose a face representation approach using multi-orientation Log-Gabor local binary pattern (MOLGLBP) for realizing face recognition under facial expressions, illuminations and partial occlusions. Log-Gabor filters with different scales (frequencies) and orientations are applied on Y, I, and Q channel image in the YIQ color space respectively. Then Log-Gabor images of different orientations at the same scale are combined to form a multi-orientation Log-Gabor image (MOLGI) and two LBP operators are applied to it. For face recognition, histogram intersection metric is utilized to measure the similarity of faces. The proposed approach is evaluated on the CurtinFaces database and experiments demonstrate that the proposed approach is effectiveness against two simultaneous variations: expression & illumination, and illumination & occlusion.

The new generation video standard, i.e., High-efficiency Video Coding (HEVC), shows a significantly improved efficiency relative to the last standard, i.e., H.264. However, the quad tree structured coding units (CUs), which are adopted in HEVC to improve compression efficiency, cause high computational complexity. In this study, a novel fast algorithm is proposed for CU partition in intra coding to reduce the computational complexity. A rough minimum depth prediction of the largest CU method and an early termination method for CU partition based on the total coding bits of the current CU are employed. Many approaches have been proposed to reduce the encoding complexity of HEVC, but these methods do not use the total coding bits of the current CU as the main basis for judgment to judge the CU complexity. Compared with the reference software HM16.6, the proposed algorithm reduces encoding time by 45% on average and achieves an approximately 1.1% increase in Bjntegaard delta bit rate and a negligible peak signal-to-noise ratio loss.

Person Re-Identification has received extensive study in the past few years and achieves impressive progress. Recent outstanding methods extract discriminative features by slicing feature maps of deep neural network into several stripes. Still there have improvement on feature fusion and metric learning strategy which can help promote slice-based methods. In this paper, we propose a novel framework that is end-to-end trainable, called Multi-level Slice-based Network (MSN), to capture features both in different levels and body parts. Our model consists of a dual-branch network architecture, one branch for global feature extraction and the other branch for local ones. Both branches process multi-level features using pyramid feature alike module. By concatenating the global and local features, distinctive features are exploited and properly compared. Also, our proposed method creatively introduces a triplet-center loss to elaborate combined loss function, which helps train the joint-learning network. By demonstrating the comprehensive experiments on the mainstream evaluation datasets including Market-1501, DukeMTMC, CUHK03, it indicates that our proposed model robustly achieves excellent performance and outperforms many of existing approaches. For example, on DukeMTMC dataset in single-query mode, we obtain a great result of Rank-1/mAP =85.9%(+1.0%)/74.2%(+4.7%).

A single-electron turnstile and electrometer circuit was fabricated on a silicon-on-insulator substrate. The turnstile, which is operated by opening and closing two metal-oxide-semiconductor field-effect transistors (MOSFETs) alternately, allows current quantization at 20 K due to single-electron transfer. Another MOSFET is placed at the drain side of the turnstile to form an electron storage island. Therefore, one-by-one electron entrance into the storage island from the turnstile can be detected as an abrupt change in the current of the electrometer, which is placed near the storage island and electrically coupled to it. The correspondence between the quantized current and the single-electron counting was confirmed.

An energy efficient modulator for an ultra-low-power (ULP) 60-GHz IEEE transmitter is presented in this paper. The modulator consists of a differential duobinary coder and a semi-digital finite-impulse-response (FIR) pulse-shaping filter. By virtue of differential duobinary coding and pulse shaping, the transceiver successfully solves the adjacent-channel-power-ratio (ACPR) issue of conventional on-off-keying (OOK) transceivers. The proposed differential duobinary code adopts an over-sampling precoder, which relaxes timing requirement and reduces power consumption. The semi-digital FIR eliminates the power hungry digital multipliers and accumulators, and improves the power efficiency through optimization of filter parameters. Fabricated in a 65nm CMOS process, this modulator occupies a core area of 0.12mm2. With a throughput of 1.7Gbps/2.6Gbps, power consumption of modulator is 24.3mW/42.8mW respectively, while satisfying the IEEE 802.11ad spectrum mask.

Serial RapidIO (SRIO) is a high-performance interconnection standard for embedded systems. Cyclic Redundancy Check (CRC) provides protection for packet transmissions and impacts the device performances. In this paper, two CRC calculation strategies, based on adjustable slicing parallelization and simplified calculators, are proposed. In the first scheme, the temporary CRC result of the previous cycle (CPre) is considered as a dependent input for the new cycle and is combined with a specific segment of packet data before slicing parallelization. In the second scheme, which can reach a higher maximum working frequency, CPre is considered as an independent input and is separated from the calculation of packet data for further parallelization. Performance comparisons based on ASIC and FPGA implementations are demonstrated to show their effectiveness. Compared with the reference designs, more than 34.8% and 13.9% of average power can be improved by the two proposed schemes at 156.25MHz in 130nm technology, respectively.

The demand for wireless traffic is increasing rapidly, which has posed huge challenges to mobile network operators (MNOs). A heterogeneous network (HetNet) framework, composed of a marcocell and femtocells, has been proved to be an effective way to cope with the fast-growing traffic demand. In this paper, we assume that both the macrocell and femtocells are owned by the same MNO, with revenue optimization as its ultimate goal. We aim to propose a pricing strategy for macro-femto HetNets with a user centric vision, namely, mobile users would have their own interest to make rational decisions on selecting between the macrocell and femtocells to maximize their individual benefit. We formulate a Stackelberg game to analyze the interactions between the MNO and users, and obtain the equilibrium solution for the Stackelberg game. Via extensive simulations, we evaluate the proposed pricing strategy in terms of its efficiency with respect to the revenue optimization.

Aperture synthesis technology represents an effective approach to millimeter-wave radiometers for high-resolution observations. However, the application of synthetic aperture imaging radiometer (SAIR) is limited by its large number of antennas, receivers and correlators, which may increase noise and cause the image distortion. To solve those problems, this letter proposes a compressive regularization imaging algorithm, called CRIA, to reconstruct images accurately via combining the sparsity and the energy functional of target space. With randomly selected visibility samples, CRIA employs l1 norm to reconstruct the target brightness temperature and l2 norm to estimate the energy functional of it simultaneously. Comparisons with other algorithms show that CRIA provides higher quality target brightness temperature images at a lower data level.

Network traffic load usually differs significantly at different times of a day due to users' different time-preference. Network congestion may happen in traffic peak times. In order to prevent this from happening, network service providers (NSPs) can either over-provision capacity for demand at peak times of the day, or use dynamic time-dependent pricing (TDP) scheme to reduce the demand at traffic peak times. Since over-provisioning network capacity is costly, many researchers have proposed TDP schemes to control congestion as well as to improve the revenue of NSPs. To the best of our knowledge, all the studies on TDP schemes consider only the monopoly or duopoly NSP case. In our previous work, the duopoly NSP case has been studied with the assumption that each NSP has complete information of quality of service (QoS) of the other NSP. In this paper, an oligopoly NSP case is studied. NSPs try to maximize their overall revenue by setting time-dependent price, while users choose NSPs by considering their own time preference, congestion status in the networks and the price set by the NSPs. The interactions among NSPs are modeled as an oligopoly Bertrand game. Firstly, assuming that each NSP has complete information of QoS of all NSPs, a unique Nash equilibrium of the game is established under the assumption that users' valuation of QoS is uniformly distributed. Secondly, the assumption of complete information of QoS of all NSPs is relaxed, and a learning algorithm is proposed for NSPs to achieve the Nash equilibrium of the game. Analytical and experimental results show that NSPs can benefit from TDP scheme, however, not only the competition effect but also the incomplete information among NSPs causes revenue loss for NSPs under the TDP scheme.

Without recourse to the Shannon-Nyquist sampling theorem, a novel information sampling (IS) concept is proposed for ultra-wideband (UWB) communications. To implement IS, a random pre-coding system architecture is designed and system performance is studied. Simulation results from one of UWB channel models show that the proposed system is effective to detect UWB signals with a low-sampling-rate analog-to-digital converter (ADC) at the receiver. Moreover, it can operate in a regime of heavy inter-symbol interference (ISI).

Nonnegative component representation (NCR) is a mid-level representation based on nonnegative matrix factorization (NMF). Recently, it has attached much attention and achieved encouraging result for action recognition. In this paper, we propose a novel hierarchical dictionary learning strategy (HDLS) for NMF to improve the performance of NCR. Considering the variability of action classes, HDLS clusters the similar classes into groups and forms a two-layer hierarchical class model. The groups in the first layer are disjoint, while in the second layer, the classes in each group are correlated. HDLS takes account of the differences between two layers and proposes to use different dictionary learning methods for this two layers, including the discriminant class-specific NMF for the first layer and the discriminant joint dictionary NMF for the second layer. The proposed approach is extensively tested on three public datasets and the experimental results demonstrate the effectiveness and superiority of NCR with HDLS for large-scale action recognition.

With the popularity of smart devices, mobile crowdsensing, in which the crowdsensing platform gathers useful data from users of smart devices, e.g., smartphones, has become a prevalent paradigm. Various incentive mechanisms have been extensively adopted for the crowdsensing platform to incentivize users of smart devices to offer sensing data. Existing works have concentrated on rewarding smart-device users for their short term effort to provide data without considering the long-term factors of smart-device users and the quality of data. Our previous work has considered the quality of data of smart-device users by incorporating the long-term reputation of smart-device users. However, our previous work only considered a quality maximization problem with budget constraints on one location. In this paper, multiple locations are considered. Stackelberg game is utilized to solve a two-stage optimization problem. In the first stage, the crowdsensing platform allocates the budget to different locations and sets price as incentives for users to maximize the total data quality. In the second stage, the users make efforts to provide data to maximize its utility. Extensive numerical simulations are conducted to evaluate proposed algorithm.

In this paper, we report the recent progress in visible light positioning and communication systems using light-emitting diodes (LEDs). Due to the wide deployment of LEDs for indoor illumination, visible light positioning (VLP) and visible light communication (VLC) using existing LEDs fixtures have attracted great attention in recent years. Here, we review our recent works on visible light positioning and communication, including image sensor-based VLP, photodetector-based VLP, integrated VLC and VLP (VLCP) systems, and heterogeneous radio frequency (RF) and VLC (RF/VLC) systems.