The future Internet-of-Things network is expected to connect billions of sensors, which incurs high latency for data aggregation. To overcome this challenge, a new technique called over-the-air function computation was recently developed enable fusion center receive desired directly. It utilizes the superposition property wireless channel realize uniform summation function. In order compensate non-uniform fading different we propose novel uniform-forcing transceiver design computation. A...

In the context of globalization, accuracy English pronunciation has become increasingly important for non-native speakers. This study aims to explore application and effectiveness deep learning techniques in correction. Through a series experimental studies, we compared with traditional teaching methods, focusing on measures such as accuracy, fluency, learner satisfaction. The found that were significantly better than methods improving achieved higher face challenges data quality model...

Neural networks, as powerful models for many difficult learning tasks, have created an increasingly heavy computational burden. More and more researchers focus on how to optimize the training time, one of difficulties is establish a general iteration time prediction model. However, existing high complexity or tedious build processes, there still space improvement in accuracy. Moreover, little systematic analysis multi-GPU which special widely used scenario. In this paper, we introduce...

Graph convolutional neural netwoks (GCNNs) have been emerged to handle graph-structured data in recent years. Most existing GCNNs are either spatial approaches working on neighborhood of each node, or spectral based graph Laplacian. Compared with spatial-based GCNNs, spectral-based capable highly exploiting structure information, but always regard graphs undiredcted. Actually, there many scenarios where the structures directed, such as social networks, citation etc. Treating undirected may...

Reinforcement Learning (RL), especially Deep (DRL), has made great progress in many areas, such as robots, video games and driving. However, sample inefficiency is a big obstacle to the widespread practical application of DRL. Inspired by decision making human brain, this problem can be solved incorporating instance based learning, i.e. episodic memory. Many memory RL algorithms have emerged recently. these either only replace parametric DRL algorithm with control or incorporate single...

Multipath TCP has attracted increasing attention as a promising technology for 5G networks. To fully utilize network interfaces on multi-homed terminals and the whole resources, MPTCP is proposed an extension of to transfer packets concurrently over multiple paths. Cross layer optimization techniques have been applied such routing path management. However, existing multipath algorithms modeling are facing challenges subflow asymmetry due heterogeneity, thus cannot handle problems...

Cloud storage attracts a lot of clients to join the paradise. For high data availability, some require their files be replicated and stored on multiple servers. Because are generally charged based redundancy level required by them, it is critical for obtain convincing evidence that all replicas correctly updated up-to-date version. In this article, we propose dynamic proof possession replication (DPDPR) scheme, which proved secure in defined security model. Our scheme shares single...

Deploying relays and/or base stations is a major means of extending the coverage wireless network. This paper presents models, analytical results, and algorithms to answer two related questions: The first where deploy extend reach from station maximum, second how many are needed any point in given area. An important use case results public safety broadband network, which deploying often crucial providing an incident scene.

The paper describes a novel ant-based distributed route algorithm (ADRA) for ad-hoc networks. ants move across the network between randomly chosen pairs of nodes; as they deposit simulated pheromones function their distance from source node, quality link, congestion encountered on journey, current nodes possess and velocity at which move. Of course, node changes by itself according to ages link evaporating. An ant selects its path each intermediate distribution node. In order accelerate...

The minimum variance distortionless response (MVDR) beamformer is very sensitive to errors, such as the finite snapshots and steering vector errors. A small mismatch can lead serious degradation its performance, especially at high signal-to-noise ratio(SNR). In this paper, a new robust MVDR proposed. Our analysis starts from physical meaning of weight formula. Based on analysis, DOA matrix constructed in single uniform linear array. Through decomposition matrix, we could obtain actual signal...

Abstract For a high level of data availability and reliability, common strategy for cloud service providers is to rely on replication, i.e. storing several replicas onto different servers. To provide users with strong guarantee that all required by them are actually stored, many multi-replica integrity auditing schemes were proposed. However, most existing solutions not resource economical since need create upload their files themselves. A solution called Mirror presented overcome the...

With the enlargement of network size and application scope mobile ad hoc networks, variety number service needed by such networks become more more. Furthermore, in this dynamic environment, different nodes offering services may enter leave at any time. Efficient timely discovery becomes a new task network. Discovery is crucial feature for usability networks. In paper, we survey researches recent years on discovery, classify existing protocols Moreover, analyze compare these protocols. Based...

The broadband satellite multimedia (BSM) architecture standardised by ETSI defines a independent service access point (SI-SAP) interface layer that separates the features of upper layers from dependant lower layers, and provides mechanism to carry IP-based protocols over these dependent layers. This enables interoperability at IP between systems different physical link technologies fully comply with SI-SAP concept. study reviews past current standardisation activities including BSM quality...

In order to reflect the actual case of users' decisions and rating patterns, solve sparsity problem traditional collaborative filtering recommendation algorithms, a trapezoid fuzzy model is proposed, which fuzzifies crisp point into number based on stat istical information users. The contains personalized users can reasonably represent user's preference degree pattern. Based this model, user similarity-based algorithm designed. We prove that extension Cosine similarity, analyze application...

Multipath transmission control protocol (MPTCP) allows TCP (transmission protocol) connections to operate across multiple paths concurrently help improve resource utilization as well connection robustness. Existing congestion algorithms for MPTCP are mainly loss-based, thus leading severe performance degradation in wireless environment and buffer bloat on network devices. In this paper, get rid of these shortcomings, we modify a based algorithm (bottleneck bandwidth round-trip propagation...

The paper presents a novel dynamic load-aware based load-balanced routing (DLBL) algorithm for ad hoc networks. DLBL considers an intermediate node load as the primary route selection metric while discovering route; this can lead to less network congestion and fewer bottlenecks. During maintaining periods, deals, by distributed mechanism, with of active routes when nodes or have their queue overload, at same time little overhead increment. When link breaks because mobility power off, etc.,...

Future communication systems might be characterized by an integration of multiple access systems, and people are expecting excellent multi-access solutions for better services. However, currently proposed in different layers unable to completely fulfill the requirements. In this paper, we put forward stream control transport protocol (SCTP) as a layer approach multi-access, propose efficient SCTP load-sharing enhancement. The evaluation performance has been investigated key scenarios with...

Decentralized federated learning (DFL) architecture enables clients to collaboratively train a shared machine model without central parameter server. However, it is difficult apply DFL multi-cell scenario due inadequate averaging and cross-cell device-to-device (D2D) communications. In this paper, we propose hierarchically decentralized (HDFL) framework that combines intra-cell D2D links between devices backhaul communications base stations. HDFL, from different cells global using periodic...