Using cloud computing, individuals can store their data on remote servers and allow access to public users through the servers. As outsourced are likely contain sensitive privacy information, they typically encrypted before uploaded cloud. This, however, significantly limits usability of due difficulty searching over data. In this paper, we address issue by developing fine-grained multi-keyword search schemes Our original contributions three-fold. First, introduce relevance scores preference...

Automatic modulation classification (AMC) lies at the core of cognitive radio and spectrum sensing. In this Letter, authors propose a novel convolutional neural network (CNN)-based AMC method with multi-feature fusion. First, signals are transformed into two image representations cyclic spectra (CS) constellation diagram (CD), respectively. Then, two-branch CNN model is developed, gradient decent strategy adopted fusion technique exploited to integrate features learned from CS CD. The...

The study introduces the novel message authentication schemes for smart meter system, where symmetric cryptography-based physical layer-assisted (PLAA) scheme and public key infrastructure- based PLAA are introduced. proposed integrate conventional layer mechanisms by taking advantage of temporal spatial uniqueness in channel responses, aiming to achieve fast while minimising packet transmission overhead. authors also verify their claims through extensive analysis simulation via comparing...

In this article we focus on the security in cognitive radio, which is one of needs for new technologies requiring spectrum bands. The PHY-layer CR more complex than a conventional wireless communication system because sensing and dynamic access mechanism. Therefore, it becomes vulnerable to be invaded. We present approaches defense against threats networks. analyze validities benefits proposed approaches, can valid not only attacks but also those upper layer.

The massive parallelism present in GPUs comes at the cost of reduced L1 and L2 cache sizes per thread, leading to serious contention problems such as thrashing. Hence, data access locality an application should be considered during thread scheduling improve execution time energy consumption. Recent works have tried use behavior regular structured applications scheduling, but difficult case irregular unstructured parallel remains explored. We PAVER , a P riority- A ware V ertex schedul ER...

In this letter, we consider the bit mapping (BM) design for regular low-density parity-check (LDPC) coded bit-interleaved modulation (BICM) schemes. We propose a check-node type distribution and develop multi-edge extrinsic information transfer (EXIT) functions LDPC BICM Based on EXIT functions, two-phase method to generate optimized practical BM in terms of lowest decoding threshold given finite length code. Simulation results show that proposed has better compared with those existing works.

In literatures, there are various constructions of frequency hopping sequence (FHS for short) sets with good Hamming correlations. Some papers employed only multiplicative groups finite fields to construct FHS sets, while other implicitly used additive construction sets. this paper, we make use both and simultaneously present a optimal The provides new family (q <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">m</sup> - 1, q...

In this letter, a new bound on partial Hamming correlation (PHC) of low-hit-zone frequency hopping (FH) sequences is derived. Then, the PHC conventional FH also deduced. Furthermore, two constructions sequence sets are presented, which strictly optimal with respect to by Zhou et al. They have parameters.

Modulation recognition is an important technology in wireless communication systems. In recent years, deep learning-based modulation algorithms, which can autonomously learn features and achieve superior performance compared with traditional have emerged. Yet, there are still certain limitations. this paper, aiming at addressing the issue of poor low signal-to-noise ratios (SNRs) inability to effectively distinguish among all types, we propose optimization scheme for based on fine-tuning...

In this paper, we propose a new multi-edge metric-constrained quasi-cyclic progressive edge-growth algorithm (MM-QC-PEGA), which is suitable for constructing both single- and multi-weighted (binary) QC low-density parity-check (LDPC) codes with arbitrary lengths, rates, circulant sizes, variable node (VN)-degree distributions. The MM-QC-PEGA able to detect all cyclic-edge-set-minimum-virtual cycles (CMVCs), as it accurately computes the metric value of each CMVC an posteriori test. addition,...

Achieving energy proportionality in data centers supporting latency-sensitive applications is challenging because of the strict Service Level Agreements. Previous works individually focus on making server proportional or reducing center network's power consumption for latency-tolerant applications. In this paper, we propose EPRONS to minimize overall center's with by trading-off network slack favor providing additional computations. We utilize linear programming model consolidate search...

Saving energy for latency-critical applications like web search can be challenging because of their strict tail latency constraints. State-of-the-art power management frameworks use Dynamic Voltage and Frequency Scaling (DVFS) Sleep states techniques to slow down the request processing finish just-in-time. However, accurately predicting compute demand a difficult. In this paper, we present Gemini, novel framework engines. Gemini has two unique features capture per query service time...

Batched sparse (BATS) code is a promising technology for reliable data transmission in error-prone environments. A BATS consists of an outer and inner code. It has been shown that well-designed critical to the performance BATS. Though great efforts have made design codes, impact scheduling (that determines when codes take place at nodes path) on still unclear. In this paper, we study joint multihop wireless networks. We first introduce new network utility from which associate with...

In order to evaluate the goodness of frequency hopping (FH) sequence design, periodic Hamming correlation function is used as an important measure. Usually, length window shorter than period chosen FH sequence, so study partial particularly important. If or set has optimal with respect bound for all window, then said be strictly optimal. this paper, we first prove that there no family size M and N over a slot q derived by Niu et al. when > <sup xmlns:mml="http://www.w3.org/1998/Math/MathML"...

Aperiodic Hamming correlation (APHC) is more accurate to measure performances of the frequency hopping (FH) communication systems than periodic correlation. A class no‐hit‐zone FH sequences (FHSs) with respect APHC which has not yet been studied previously, studied. bound on FHSs established. Further, a constructed, optimal new bound.

Outsourcing heavy computational tasks to remote cloud server, which accordingly significantly reduce the burden at end hosts, represents an effective and practical approach towards extensive scalable mobile applications has drawn increasing attention in recent years. However, due limited processing power of hosts yet keen privacy concerns on outsourced data, it is vital ensure both efficiency security outsourcing computation computing. In this paper, we address issue by developing a publicly...

This paper is concerned with a machine learning approach to cancel the interference for cognitive Internet of Things (C-IoT) in concurrent spectrum access (CSA) model, where C-IoT system noncooperative and has very limited knowledge on interference. Our transceiver design uses an iterative processing structure, which consists linear estimator, demodulation-and-decoding module, clustering module. In we employ modified expectation-maximization (EM)-based algorithms estimate under modulation...

Power management in data centers is challenging because of fluctuating workloads and strict task completion time requirements. Recent resource provisioning systems, such as Borg RC-Informed, pack tasks on servers to save power. However, current power optimization frameworks based packing leave very little headroom for spikes, the times are compromised. In this paper, we design Goldilocks, a novel system optimizing both by allocating groups. Tasks hosted containers grouped together running...

Abstract In this paper, we address the problem of heterogeneous quality‐of‐service (QoS) provisioning in IEEE 802.16‐based wireless mesh networks. Our objective is to provide a novel service differentiation scheme based on distributed scheduling mode, thereby offering QoS guarantees. We first devise for traffic class differentiation, which can be implemented at medium access control layer. then present QoS‐aware fair packet (QFPS) algorithm fulfill provisioning. With QFPS, flow with urgent...

Cross-layer design has been widely used in wireless sensor networks, especially to improve the network lifetime, as can be seen literature. In this paper, a cross-layer solution is combined transmission advertisement scheme slotted ALOHA-based throughput and lifetime. This medium access method chosen because it does not add protocol information transmitted with data bits, reducing overhead when compared other methods. Finally, combination of proven increase by more than 10% double

We propose a new low-density parity-check coded scheme for two-user Gaussian multiple access channels (GMACs) with equal rate and average power constraints. In the proposed scheme, we divide code word into two parts length transmit them by using BPSK signals levels. Two users use levels alternatively to avoid ambiguity problem in GMACs. optimize ratio between minimize required signal-to-noise given rate. also design joint decoding algorithm our scheme. Numerical results show that has better...

In this article, a modified mutual information maximization (InfoMax) framework, named channel capacity (CapMax), is proposed and applied to learn informative representations for dynamic networks with time-varying topology and/or time-evolving node attributes. The CapMax based on the network theory multiuser communication, where representation model treated as multiaccess communication memory feedback. Without requirements of backbone structure, learning objective our maximizing capacity,...