Communications and Networking

For mobile devices (MDs) and Internet of Things (IoT) devices with limited computing capacity and battery, offloading part of tasks to the mobile edge computing (MEC) server is attractive. In this paper, we propose a joint partial computation offloading and pricing scheme in a multi-user MEC system. Firstly, we establish MD’s cost model and MEC server’s revenue model in terms of money. Secondly, we investigate MD’s cost minimization partial offloading strategy to jointly control MD’s task allocation, local CPU frequency and the amount of computational resource blocks (CRBs) requested. Finally, we formulate the revenue maximization problem for MEC server with limited computing capacity, a heuristic algorithm is proposed for MEC server to find the optimal service price. Numerical results verify the effectiveness of our proposed scheme in cost saving and pricing.

By applying caching resource at the remote radio heads (RRHs), the fog radio access network (Fog-RAN) has been considered as an promising wireless architecture in the future network to reduce the transmission delay and release the heavy burden of backhaul link for huge data delivery. In this paper, we propose to use the Fog-RAN to assist the data transmission in the high-speed railway scenario. In specific, we investigate the dynamic resource allocation in high-speed railway Fog-RAN systems by considering the delay constraint. The instantaneous power allocation at the RRHs and the instantaneous content delivery rate over the backhaul links are jointly optimized with an aim to minimize the total power consumed at the RRHs and over the backhaul links. An alternating optimization (AO) approach is used to find solutions of the instantaneous power and instantaneous content delivery rate in two separate subproblems. The closed-form solutions are derived in two subproblems under certain special conditions. Simulation results demonstrate that the proposed dynamic resource allocation is significantly superior to the constant resource allocation scheme.

With the rapid development of the mobile internet, many emerging compute-intensive and data-intensive tasks are extremely sensitive to latency and cannot be implemented on mobile devices (MDs). To solve this problem, mobile edge computing (MEC) appears to be a promising solution. In this paper, we propose a distributed task splitting and offloading algorithm (DSOA) for the scenario of multi-device and multi-MEC servers in ultra-dense networks (UDN). In the proposed scheme, the MDs can perform their tasks locally or offload suitable percentage of tasks to the MEC server. The optimization goal is to minimize the overall task computation time. Since the MDs are selfish, we propose a game theory approach to achieve optimal global computation time. Finally, the numerical simulation results verify that the algorithm can effectively reduce global computation time.

Incorporating visible light communication (VLC) with existing radio frequency (RF) access techniques has received widespread concern to enhance network coverage/capacity. This paper focuses on the joint downlink resource allocation (RA) in a hybrid VLC-RF network. The problem is formulated as utility maximization by jointly adjusting downlink sub-channel allocation. A Evolution Computation (EC) based centralized algorithm is developed to solve the problem. To reduce computation complexity, the algorithm is decoupled into two sub-steps. First, users are assigned to different VLC access points and the allocation is initialized in a proportional fair (PF) like method. Second, EC search procedures are iteratively operated until optimality. Through simulation, the algorithm outperforms classic PF and Round Robin RA methods in terms of throughput and user fairness.

Mobile Edge Computing (MEC) has become the most likely network architecture to solve the problems of mobile devices in terms of resource storage, computing performance and energy efficiency. In this paper, we first model the MEC system with the exploitation of mobility prediction. Considering the user’s mobility, the deadline constraint and the limited resources in MEC servers, we propose a deep reinforcement learning approach named deep deterministic policy gradient (DDPG) to learn the power allocation policies for MEC servers users. Then, the aim of the policy is to minimize the overall cost of the MEC system. Finally, simulation results are illustrated that our proposed algorithm achieves performance gains.

Analytical EDF Priority schedulers are not common in Mobile Ad-hoc Networks (MANETs). Some researchers like Abhaya et al. have proposed a classical preemptive Earliest Deadline First (EDF) scheduler. The goal of this EDF scheduler was to favor higher priority packets thereby reducing their waiting times. Accordingly, favoring higher priority queues end up increasing the waiting times of lower priority queues. We improve Abhaya’s approach and adopt it to the MANETs environment. We numerically study the performance of the Adopted and Improved Adopted Abhaya Earliest Deadline First (IEDF) models for different packet queues. Our analytical results show that the IEDF model shortens the waiting times of packets of the different queues at various system loads in comparison to the Adopted Abhaya EDF model.

Edge computing is a new network model providing low-latency service with low bandwidth cost for the users by nearby edge servers. Due to the limited computational capacity of edge servers and devices, some edge servers need to offload some tasks to other servers in the edge network. Although offloading task to other edge servers may improve the service quality, the offloading process will be charged by the operator. In this paper, the goal is to determine the task offloading decisions of all the edge servers in the network. A model is designed with different types of cost in edge computing, where the overall cost of the system reflects the performance of the network. We formulate a cost minimization problem which is NP-hard. To solve the NP-hard problem, we propose a Joint Collaborative Task Offloading algorithm by adopting the optimization process in nearby edge servers. In our algorithm, an edge server can only offload its tasks to other edge servers within a neighborhood range. Based on the real-world data set, an adequate range is determined for the edge computing network. In cases of different density of tasks, the evaluations demonstrate that our algorithm has a good performance in term of overall cost, which outperforms an algorithm without considering the influence of neighborhood range.

Compared with uncoded caching, coded caching (CC) that exploits accumulated cache size and distributes fractions of a file in different base stations (BSs), can significantly reduce delays and backhaul transmissions. Small base stations (SBSs) with both cache and energy harvesting (EH) ability have attracted extensive attention in recent years. Moreover, renewable energy (RE) also called green energy can be shared between SBSs via the Smart Grid (SG). This paper investigates CC, resource allocation (RA) and energy cooperation (EC) in cache-enabled energy harvesting (EH) heterogeneous networks (HetNets). We formulate the joint optimization problem, aims at minimizing the conventional grid energy consumption while satisfying quality of service (QoS) requirements of users. Simulation results demonstrate the considerable reduction in conventional grid energy consumption compared with other benchmarks.

With the emergence of diverse computation-intensive mobile applications (such as virtual reality), demands for data processing from users are rapidly increasing in mobile edge computing (MEC). However, existing mobile edge servers (MES) are susceptible to propagation delays and loss and fail to provide timely and efficient services. Facing this problem, we focus on applying unmanned aerial vehicles (UAVs) equipped with computing resources to provide mobile edge computing offload services for users. UAV as an MES can guarantee low propagation delay and high reliability due to its maneuverability and short-distance line-of-sight communications. In this paper, we study a joint computing offloading problem consideration of user equipments, ground base stations and aerial UAVs. The system provides two offloading methods. The first offloading method is the air-offloading, where a user equipment can offload computing tasks to UAV-enabled MEC servers. The second offloading method is ground-offloading, where a user equipment can offload computing tasks to existing MESs. The task-aware optimization offloading scheme is proposed and it selects local execution or an offloading method based on the latency and energy constraints. Simulation results show that our proposed offloading selection scheme outperforms benchmark schemes. The results demonstrate that the proposed schemes improve quality of service (QoS) and have low task block rate under latency and energy constraints.

Smart grid achieves optimal management of the entire power system operation by constant monitoring and rapid demand response (DR) for power supply-demand balance. Constantly monitoring the system state realized by Wide Area Measurement Systems (WAMS) provides a global view of the power grid. With a global view of the grid, Wide Area Control (WAC) generated DR command to improve the stability of power systems. When the regular monitoring data flow and the sudden DR data coexist, the suddenness of the demand response may result in delay or loss of the data packet due to uneven resource allocation when the network communication resources are limited, thereby affecting the accuracy of the power system state estimation. To solve this problem, this paper proposes a burst traffic perception weighted round robin algorithm (BTAWRR). The proposed algorithm defines the weight of the cyclic scheduling according to the periodicity of the monitoring data and the suddenness of the demand response. Then it adopts the iterative cyclic scheduling to adjust the transmission of data packets in time by adaptively sensing the changes of the traffic flow. The simulation results show that the proposed algorithm can effectively reduce the scheduling delay and packet loss rate when the two data coexist, and improve the throughput, which is beneficial to ensure the stability of the smart grid.

In this paper, we consider a cooperative computing system which consists of a number of mobile edge computing (MEC) servers deployed with convolutional neural network (CNN) model, a remote mobile cloud computing (MCC) server deployed with CNN model and a number of mobile devices (MDs). We assume that each MD has a computation task and is allowed to offload its task to one MEC server where the CNN model with various layers is applied to conduct task execution, and one MEC server can accept multiple tasks of MDs. To enable the cooperative between the MEC servers and the MCC server, we assume that the task of MD which has been processed partially by the CNN model of the MEC server will be sent to CNN model of the MCC server for further processing. We study the joint task offloading, CNN layer scheduling and resource allocation problem. By stressing the importance of task execution latency, the joint optimization problem is formulated as an overall task latency minimization problem. As the original optimization problem is NP hard, which cannot be solved conveniently, we transform it into three subproblems, i.e., CNN layer scheduling subproblem, task offloading subproblem and resource allocation subproblem, and solve the three subproblems by means of extensive search algorithm, reformulation-linearization-technique (RLT) and Lagrangian dual method, respectively. Numerical results demonstrate the effectiveness of the proposed algorithm.

At present, Multi-platform Avionics System (MPA) has been widely used. The existing adaptive scheduling algorithm based on Sliding-Scheduled Tenant (SST) simulates and verifies the resource management and task scheduling of MPA, and analyzes the task requirements of MPA. However, due to the shortcomings of SST algorithm in considering energy consumption and other aspects, it reduces the task acceptance rate, and does not consider the limitations of sensors and priorities, which makes al algorithm unable to meet the requirements of avionics system. This paper proposes a method of system resource selection, which considers the energy consumption, sensor and priority limit, so as to improve the acceptance rate of tasks, improve the acceptance rate of high priority, and get a scheduling algorithm with high acceptance rate of tasks. Finally, through the comprehensive analysis of the experimental results and experimental results in different scenes, it is shown that the algorithm proposed in this paper outperforms the existing algorithm in terms of the acceptance rate.

In the process of missions, how to transmit messages to the destination node quickly is a crucial issue for UAVs. Some existing methods show bad effects such as low delivery ratio, long delay, large average hop count, and high ping-pong effect ratio, thus this paper proposes a new algorithm. By considering the position of all UAVs at each moment, UAVs can obtain optimal message transmission, thus get the optimal path for the message to reach the destination node. After doing simulation experiments with the existing algorithms as DTNgeo, DTNclose and DTNload, the DPTM algorithm is superior to those in terms of delivery ratio, delay, hop count and ping-pong effect ratio.

Based on the theory of classical electrodynamics and quantum mechanics, we quantitatively deduce microwave carrying Orbital Angular Momentum (OAM) radiated from the moving free electron beams on different closed-curved trajectories. It shows that the non-relativistic free electrons can also transit quantized OAM to the microwave in addition to the relativistic cyclotron electrons in the magnetic field. This work indicates the effective way to construct the antennas to generate high OAM modes of the microwave by multi-electron radiation.

This paper investigates the semi-persistent scheduling (SPS) strategy for enhanced vehicle-to-everything (eV2X) services, which aims to meet the low latency and high reliability (LLHR) demands. To increase available spectrum and improve resource utilization, millimeter wave (mmWave) and non-orthogonal multiple access (NOMA) are considered. We first formulate the optimization problem of scheduling and resource allocation to minimize the SPS period. To solve this problem, the LLHR power control algorithm is proposed to provide evaluation indicators for user scheduling. Then, the beam division and user clustering algorithm is designed to reduce the complexity of the matching between users and resource blocks. After that, the matching problem with peer effects is solved by the proposed union-based matching algorithm. Complexity analysis is presented, and simulation results show that the scheduling period of eV2X systems can be improved by the proposed SPS strategy compared with the conventional mmWave SPS schemes.

Aiming at the requirement of underwater information security transmission, the security of encryption key generation and distribution in underwater acoustic communication is concerned. Key generation technology based on underwater acoustic channel (UAC) estimation can improve the security and real-time generation of encryption keys. In this paper, the idea of estimating the multipath structure of UAC is to retrieve the arrival signal by acquiring the parameters of larger energy Eigen-ray from real arrival signal, and to eliminate the arrival signal of larger energy Eigen-ray path from the real signal through signal cancellation, so as to eliminate the influence of side lobes of larger energy signal to arrival signals of other Eigen-ray path, to improve the estimation performance of multipath structure in underwater acoustic channel. The simulation and experimental results show that the improved algorithm can estimate the multipath structure of underwater acoustic channel more accurately and provide support for the subsequent underwater information security transmission.

With the rapid development of wireless communications technology, the problem of scarcity of spectrum resources is becoming serious. Cognitive radio (CR) which is an effective technology to improve the utilization of spectrum resources is getting more and more attention. Spectrum sensing is a key technology in cognitive radio. Wideband spectrum sensing (WBSS) can help secondary users (SUs) find more spectrum holes. However, for the traditional energy detection (ED) algorithm, when the signal-to-noise ratio (SNR) of the primary user (PU) is low, the detection performance is extremely poor owing to the single frequency point detection method. Therefore, the concept of spectrum correlation is proposed. Spectrum correlation algorithm uses the detection window to realize joint detection of multiple frequency points which can improve performance. This paper focuses on how to make the best of spectrum correlation to ensure the detection performance for low SNR signals. We propose an adaptive detection window (ADW) method, whose detection window is adaptively selected based on the estimated SNR of signal. The method can be directly used for wideband spectrum sensing when the approximate position of each signal and its estimated SNR are known. In this context, to show the robustness of the ADW method, a simulation of the sensitivity of the ADW method to the SNR estimation error is performed. Meanwhile, simulations of methods comparison demonstrate that the proposed ADW method outperforms the commonly used iterative energy detection method, frequency correlation methods and histogram-based segmentation method by far.

Audio watermarking is an information hiding technology which is widely used in copyright protection and information security. This paper proposes a novel audio watermarking scheme based on spread spectrum and non-uniform quantization. The watermarks are embedded by modifying the quantization coefficients. The proposed algorithm utilizes the characteristics of non-uniform quantization to adopt different quantized signal-to-noise ratios for the low-frequency and high-frequency parts of the audio signal, thus improving the robustness of the technology while ensuring the sound quality is not damaged. Compared with the existing audio watermarking methods, the proposed scheme is especially robust against additive white Gaussian noise (AWGN). Experimental analysis shows that the proposed method provides high audio quality and excellent capability to withstand various noise attacks particularly in AWGN.

With the rapid development of cellular networks, demand for real-time stream is increasing dramatically. How to guarantee better quality-of-service (QoS) of real-time stream services under limited resources becomes an increasingly important issue. This paper proposes a delay based hierarchical downlink scheduling (DBS) algorithm for real-time stream in cellular networks. The hierarchical scheduler is divided into two levels. The upper level scheduler offers a prediction of the number of data bits that each real-time stream needs to guarantee the QoS. The lower level scheduler classifies all real-time streams into grade A, B and C according to Head of Line Delay and allocates resources to streams according to their different grades. The simulation results show that our algorithm performs better than other real-time schedulers, such as frame level scheduler (FLS), Modified Largest Weight Delay First (M-LWDF), EXP/PF and EXP-LOG in the aspects of delay and throughput.

Physical broadcast channel (PBCH) in 5G new radio (NR) systems transmits system informations required for the user equipment (UE) to access the cell. In the long term evolution (LTE) system, multiple PBCHs are usually combined to improve demodulation performances in the case of poor channel conditions. However, in 5G NR systems, the payload of PBCH includes the system frame number and the payloads of multiple frames are not exactly consistent. Hence, it is impossible to adopt the same combination approach as that in LTE. In this paper, there proposes a method to solve the problem of combining multiple PBCH blocks in 5G NR systems. The main idea is to convert log likelihood ratios (LLRs) of all transmitted PBCH blocks into that of the first block and accumulate all LLRs at the receiving end. Then, an improved combination algorithm is considered to reduce the complexity. The simulation results show that the proposed combination algorithm can correctly combine multiple PBCH blocks. Besides, the improved combination algorithm with sort can also reduce the complexity.

Airborne Tactical Network is a promising and special mobile Ad hoc network, connecting the ground stations and all kinds of flying combat aircrafts on battlefield through tactical data links. Designing a low delay, large capacity, high flexibility, strong scalability, and multi-priority traffic differentiated medium access control (MAC) protocol is a great challenge in the researches and applications of ATNs. In order to overcome the disadvantages in IEEE 802.11 Distributed Coordination Function (DCF) and Time Division Multiple Access (TDMA) protocols, we present a channel threshold based multiple access (CTMA) protocol for ATNs in this paper. The CTMA protocol is a novel random contention type of MAC protocols, and it can differentiate multiple priority services, and utilize multi-channel resource based on channel awareness. We intensively describe the channel occupancy statistic mechanism, multi-queueing and scheduling mechanism of multi-priority services, and channel threshold based admission control mechanism involved in the protocol. We further derive the channel threshold of each priority service, the expressions of the successful transmission probability and mean delay mathematically. Simulation results show that the CTMA protocol can differentiate services for different priorities in ATNs according to the real-time channel state, and provide effective QoS guarantee for transmissions of various information.

Low Earth Orbit (LEO) Satellite Networks (SN) offers communication services with low delay, low overhead, and flexible networking. As service types and traffic demands increase, the multi-service routing algorithms play an important role in ensuring users’ Quality of Service (QoS) requirements in LEO-SN. However, the multi-service routing algorithm only considers the link QoS information, ignoring the uneven distribution of ground users, causing satellite link or node congestion, increasing the packet transmission delay, and packet loss rate. In order to solve the above problems, we propose a Multi-Service Routing with Guaranteed Load Balancing (MSR-GLB) algorithm which balances the network load while satisfying multi-service QoS requirements. Firstly, the Geographic Location Information Factors (GLIF) are defined to balance the network load by scheduling the ISLs with lower loads. Then, the optimization objective function is constructed by considering delay, remaining bandwidth, packet loss rate, and GLIF in order to characterize the routing problems caused by multi-service and load balancing. Following this, we propose an MSR-GLB algorithm that includes the state transition rule and the pheromone update rule. Among them, the state transition rule is based on QoS information and link GLIF, and the pheromone update rule has the characteristics of positive and negative feedback mechanism. The simulation results show that the MSR-GLB algorithm can well meet the QoS requirements of different services, balance the network load compared to Cross-layer design and Ant-colony optimization based Load-balancing routing algorithm in LEO Satellite Network (CAL-LSN) and Multi-service On-demand Routing (MOR) algorithm.

The Electro-Magnetic (EM) waves with Orbital Angular Momentum (OAM) can achieve high spectral efficiency by multiplexing different OAM modes. Different modes are mapped to the frequency shifts in the secondary frequency domain at the receiving end, in order to effectively identify the OAM modes received in partial phase plane. The traditional method requires high-speed acquisition equipment in the process of receiving Radio Frequency (RF) signals directly and its hardware cost is high. Even if analog devices are used for down-conversion to Intermediate Frequency (IF) sampling, the IF bandwidth limits the transmission rate. However, Compressive Sensing (CS) can break the Nyquist restriction by random observation, and is expected to realize the detection and identification of different OAM modes at a lower sampling rate, so that the cost is low. Therefore, this paper proposes an OAM mode identification method based on CS. At the same time, the random sampling is carried out based on the existing hardware device, i.e. Analog-to-Information Converter (AIC), to realize the OAM modes identification with the low sampling rate. The simulation results verify the correctness and effectiveness of the method.

The error propagation problem in incremental freezing (IF) hybrid automatic repeat request (HARQ) scheme using Polar codes is studied. We propose two IF HARQ schemes using polar codes, namely the cyclic redundancy check (CRC)-aided IF HARQ scheme and the cumulative-path-metrics-based IF HARQ scheme. In the CRC-aided IF HARQ scheme, several CRC bits are added to each transmitted block. Using these CRC bits, the proposed IF HARQ scheme and the Chase Combining HARQ scheme can be combined to achieve a better error correction performance in the cost of a larger decoding delay. In the cumulative-path-metrics-based IF HARQ scheme, the successive joint decoder maintains multiple possible paths simultaneously, and the cumulative path metrics is used to represent the reliability of each surviving path in the decoding process. Moreover, a modified path splitting reduced successive cancellation list (SCL) decoding algorithm is presented to reduce the computational complexity and the memory requirement of cumulative-path-metrics-based IF HARQ scheme. Simulation results show that, using the Polar code constructed under long block length and high block error rate, the CRC-aided IF HARQ scheme has a higher system throughput. With the Polar code constructed under short block length and low block error rate, the cumulative-path-metrics-based IF HARQ scheme has a higher system throughput. In both situations, the system block error rate of the CRC-aided IF HARQ scheme performs well.

Intelligent reflecting surface (IRS) is currently adopted by massive multiple-input multiple-output (MIMO) systems as a new expansion scheme. It effectively copes with the increasing cost and energy consumption. In this paper, we concentrate on an IRS-assisted MIMO system, in which the base station, IRS and user are all equipped with multiple antennas. We first give the upper bound of the ergodic capacity of the system. Then we maximize this upper bound and obtain the sub-optimal phase shifts of IRS by applying semi-definite relax and Gaussian random methods. Numerical results shows the advantage of the proposed solution and the performance increase brought by multiple antennas.

In this paper, we study an unmanned aerial vehicle (UAV) assisted Fog-RAN network where an UAV perform as mobile remote radio head (RRH) to help base station forwards signals to the multiple users in the downlink transmissions, and a dedicated ground station (GS) acts as baseband unit (BBU) pool. To achieve fairness among users, we minimize the maximum transmission delay for all terrestrial users in downlink communication by jointly optimizing the user scheduling and the UAVs trajectory. Since the formulation problem is an integer non-convex optimization problem, we propose an effective iterative algorithm to find efficient solutions by using Majorize Minimization (MM) algorithm. We also confirm the convergence of our proposed algorithm. Numerical results indicate that the proposed algorithm can significantly reduces transmission delay compared to circular trajectories and fixed base station solutions.

Due to the characteristics of low latency and proximity discovery, D2D communication is considered to have an inherent advantage in supporting Internet-of-Vehicles (IoV) service. In this paper, considering that vehicular users can detect neighbor nodes in adjacent areas which are able to maintain high reliable communication with themselves, a novel design of D2D-Clustering algorithm is proposed in order to improve the QoS of users. The algorithm uses undirected graph to describe the neighborhood relationship between users. And the undirected graph is continuously simplified by multi-round traversal of vehicular users until user clustering is completed. Simulation results prove the validity of the proposed algorithm, pointing out that it helps reduce the energy consumption of the whole system.

Due to its good extendibility and robustness, Ad hoc network has been widely used in various aspects. However, its performance is restricted by the mobility, limited bandwidth and centerless architecture. In order to improve the performance of Ad hoc network, this paper proposes a cross-layer Hexagonal Clustering, Position and Time slot based (HCPT) protocol, in which clusters and time slots are divided according to the geographical location. And furthermore, an effective algorithm to find routes through the geographical locations of cluster heads is proposed, which can greatly reduce message collisions and reduce network overhead. By doing simulations in the Network Simulator 2 (NS2) software, we found that the HCPT protocol shows better performance in network topology discovery compared with the Optimized Link State Routing (OLSR) protocol. Simulation results also show that the proposed scheme outperforms the standard OLSR and the improved OLSR algorithm, which is proposed by N. Harrag, in terms of routing overhead and packet delivery ratio.

Ultra-Dense Networks (UDN) can greatly meet the demand for explosively growing data traffic via deploying small cells (SCs) densely. However, the SCs densification causes higher energy consumption and more severe inter-cell interference (ICI). The SC on/off control is one of the effective ways to solve above problems, but the challenge is to maintain network coverage while avoiding degradation of the quality of service (QoS) of user equipment (UEs). In this paper, we formulate energy efficiency (EE) optimization problem in stochastic geometry-based network and take into consideration the QoS of UEs and ICI to maximize the EE. The solution is obtained by dividing the problem into SCs clustering and intra-cluster SC on/off control. We first use an improved K-means clustering algorithm to divide the dense SCs into disjoint clusters according to the distance and density of SCs. Then, within each cluster, selecting a SC as the cluster head (CH) is responsible for performing SC on/off operations under taking minimum rate of UEs and ICI as constraints. In addition, a heuristic search algorithm (HSA) is proposed for the intra-cluster SC on/off control. Simulation results demonstrate that the proposed scheme can effectively improve the network energy efficiency and suppress interference.

For peer-to-peer (P2P) video-on-demand (VoD) streaming, this paper proposes a new P2P VoD scheme based on Dynamic Adaptive Streaming over HTTP (DASH), called P2P-DASH VoD scheme. The scheme takes advantage of both the scalability and low cost properties of P2P technology and the dynamic self-adaptation of DASH. In the proposed scheme, a multi-overlay architecture is constructed, and a DASH streaming rate control approach is proposed. The multi-overlay architecture integrates the power-law ring overlay structure and the Fibonacci ring overlay structure. Peers can search the target video segments based on the power-law ring overlay structure or the Fibonacci ring overlay structure according to the search distance. The integrated overlay structure can reduce the jump latency caused by VCR operations and improve the smoothness of playback. Furthermore, the DASH streaming rate control approach is proposed to combine DASH in P2P VoD Streaming. The DASH streaming rate control approach considers four adaptive factors (on-time arrival rate of segment, peer’s available buffer length, current overlay available bandwidth and current overlay upload bandwidth utilization). Through simulations, we demonstrate that the proposed P2P-DASH VoD scheme has short jump latency, high playback fluency and the satisfaction of users.

Cell association in heterogeneous cellular networks is a significant research issue, but existing schemes mainly optimize a single objective and could not solve such a problem with a generic utility function in polynomial time. This paper proposes a cell association scheme for generic optimization objectives with polynomial-time complexity, which employs a virtual base station method to transform it into a 2-dimensional assignment problem solved by Hungarian algorithm. Based on this scheme, a framework for the tradeoff among multiple optimization objectives is designed. This framework jointly considers spectral efficiency and load balancing, designs a weight factor to adjust their impacts on the optimization, and uses an experience pool to store the relationship between performance demands and corresponding weight factor values. For an instantaneous cell association decision in a given network scenario, the association results are obtained as soon as the corresponding factor value is taken from the pool and the Hungarian algorithm is called for the matching. Compared with existing schemes, our proposal achieves a better tradeoff between system capacity and UE fairness with an extremely low time cost.

An energy harvesting wiretap channel model is considered in which the sender is an energy harvesting node. It is assumed that at each time slot only information about the current state of the sending node is available. In order to find an effective power allocation strategy to maximize secrecy rate, we put forward a deep Q network (DQN) scheme. First, we analyze the constraints of the system and the issue of maximizing the secrecy rate. Next, the power allocation problem is formulated as a Markov Decision Process (MDP) with unknown transition probabilities. In order to solve the continuous state space problem that traditional Q learning algorithms cannot handle, we apply neural networks to approximate the value function. Finally, an online joint resource power allocation algorithm based on DQN is presented. Simulation results show that the proposed algorithm can effectively improve the secrecy rate of the model.

The ZigBee Alliance Lab proposes the concept of ZigBee-WiFi network. ZigBee-WiFi network has a broad development space when combined with the advantages of ZigBee and WiFi. However, since ZigBee and WiFi are heterogeneous in various aspects, it is necessary to find a way to interconnect the two networks. The traditional approach is to design dedicated hardware. Since the physical layer functions and part of MAC layer functions in the hardware are fixed, this method cannot adapt to the new physical layer and signal processing algorithms. Software Defined Radio (SDR) is an emerging and flexible method of transferring signal processing components from dedicated hardware to a combination of software and general purpose processors. In this paper, we use SDR in conjunction with the Universal Software Radio Peripheral (USRP) to build a flexible and universal ZigBee-WiFi gateway for interconnecting heterogeneous ZigBee and WiFi networks. The gateway has the ability to simultaneously receive and demodulate ZigBee packets, create and transmit WiFi data frames. A comprehensive performance test confirmed that the built gateway can well interconnect heterogeneous ZigBee and WiFi networks. And the built gateway provides a reference prototype for the interconnection research of heterogeneous networks.

Due to its flexible and convenient networking, Ad hoc net- works have been used in more and more scenarios. But, the features of mobility, constantly changing topologies and centerless architecture limit its applications. In order to improve the performance of Ad hoc, this paper proposes a cross-layer protocol for mobile Ad hoc network based on Hexagonal Clustering and Hybrid MAC Access (HCHMA) approach. Through the clustering algorithm, cluster heads are selected to form a backbone network for route discovery and establishment. And the MAC layer uses two different access mechanisms to ensure efficient transmission of routing packets and data packets. Benefiting from the above approaches, network overhead is greatly reduced and the through- put is improved. By doing simulations in the network simulator 2 (NS2) software, the HCHMA protocol shows better packet delivery rate, higher throughput and lower end-to-end delay compared with the Ad hoc On- demand Distance Vector Routing (AODV) protocol and the Optimized Link State Routing (OLSR) protocol.

Software Defined Networking (SDN) is a new paradigm where the forward plan is decoupled from the control plan. The controller is a central program that tells the switches and routers how to react to the incoming flows and different network changes. The placement of the controllers considering different metrics becomes a challenge in SDN WAN. In this paper, we study the controller placement problem in terms of propagation delay and load balancing. An extended K-means algorithm is introduced to partition the network into several subnetworks and place the controllers in nodes that minimize the network delay. Then a load balance index is calculated to check the effectiveness of the load balancing considering a metric $$ \varvec{\beta} $$ as the load difference between controllers. The result analysis shows that a trade off should be done between the delay and load balancing depending on the priority of the network and no optimal case can be found that minimize both of the metrics at the same time.

This paper investigates the security problem of consensus-based distributed system under false data injection attacks (FDIAs). Since the injected false data will spread to the whole network through data exchange between neighbor nodes, and result in continuing effect on the system performance, it is significant to study the impact of the attack. In this paper, we consider two attack models according to the property of the injection data, the deterministic attack and the stochastic attack. Then, the necessary and sufficient condition for the convergence of distributed system under the attack are derived, and the attack feature making the system unable to converge is provided. Moreover, the convergence result under resource-limited attack is deviated. On the other hand, the statistical properties of the convergence performance under zero-mean and non-zero-mean stochastic attacks are analyzed, respectively. Simulation results illustrate the effects caused by FDIAs on the convergence performance of distributed system.

The observation data of Argo profile floats are very crucial for long-term climate change and natural variability, which reflect three-dimensional distribution of temperature and salinity in the sea. In order to solve the anomalies in the profile caused by uncertainties factors, this paper proposes a novel anomaly detection method for Argo profile floats using an improved trajectory clustering method to discriminate normal and abnormal. The proposed algorithm partitions Argo data into a set of line segments, and then clusters line segments to get rid of noisy data, finally recovers the line segments to the raw data accordingly. As a result, the proposed oceanic anomaly detection method subtly converts the sequence data into line segments for anomaly detection, which considers both positional relationship and trend of data source. Extensive experiments on real dataset from Argo floats verify that our method has better results under different conditions compared to existing methods such as LOF and DBSCAN.

In recent years, the medical equipment and related information systems show the characteristics of mobility, networking, intelligence. At the same time, security incidents caused by medical equipment emerge in an endless stream, which brings a huge threat to the information security of users and causes serious harm. Most medical devices use open source protocol library, which brings great security risks to the digitalization and informatization of medical devices. Therefore, in the face of growing security threats and challenges, it is urgent to study the security of medical equipment. In this paper, the vulnerability mining of DICOM was studied, the most commonly used communication standard for high-performance medical devices, and a vulnerability mining model based on Fuzzing technology was proposed. This model constructed a vulnerability mining environment by simulating PACS system, and implemented a prototype system DICOM-Fuzzer. The system includes initialization, test case generation and other modules, which can complete large-scale automatic testing and exception monitoring. Then, three different versions of the open source library were selected to test the 1000 test cases generated respectively. It was found that when the received file data was greater than 7080 lines, the overflow would occur, resulting in the denial of service of the system. Finally, the security suggestions and repair measures were put forward, and the future research was described.

This paper studies physical layer security of a three node multicarrier network with a source node, a destination node and a full-duplex proactive eavesdropper who sends jamming signals for improving its eavesdropping performance. The problem of transmit power allocation for minimizing the average secrecy outage probability on all subcarriers is investigated under the assumptions that the channel state information (CSI) related to the eavesdropper is perfectly known and only channel distribution information (CDI) is known. Algorithms are proposed for the optimization problem and are shown to greatly outperform the benchmark algorithms.

Spoofing represents a significant threat to the integrity of applications relying on Global Navigation Satellite System (GNSS). A spoofer transmits counterfeit satellite signals to deceive the operation of a receiver. As multipath and spoofing signals have similar signal structures, Signal Quality Monitoring (SQM) techniques, originally designed for multipath detection, were identified to be useful for spoofing detection. Recently, a moving variance (MV) based SQM method was developed to improve the performance of raw SQM metrics. However, the main problem with implementing the MV-based SQM technique is differentiating the spoofing attack from multipath. This work presents a two-dimensional detection method using carrier power and moving variance to improve detection performance. Besides, false alarms caused by multipath are avoided by the two-dimensional threshold. A dataset called Texas Spoofing Test Battery and a multipath scenario from Osaka were employed to evaluate the performance of the proposed algorithm.

With the emergence of service port obfuscation and abuse, malicious services can hide their communication behaviors in large-scale normal SSL/TLS traffic easily. Therefore, it is of great significance to get the complete view of SSL/TLS service ports and understand the potential threat of SSL/TLS usage. In this paper, we conduct a comprehensive analysis of the SSL/TLS service port by carrying out a large-scale passive measurement based on two ISP-level networks with a total bandwidth of up to 100 Gbps for over one year. Specifically, we first investigate the overall SSL/TLS service port view and uncover that the actual usage of port is in a state of confusion. At the same time, through in-depth analysis of specific well-known ports which are used by SSL/TLS, it is revealed that the well-known ports could be exploited by malicious SSL/TLS services easily. Then, we dig into some specific certificates to explore their ports behavior and discover that the self-signed certificates and EV certificates are in sorry state. Meanwhile, we uncover practices that may be exploited by malicious services, and reveal the potential threats or vulnerabilities in SSL/TLS service ports. We believe that the work will be beneficial to both SSL/TLS and web security in the future.

In this paper, we have studied the secrecy precoder design problem for a k-user multiple-input multiple-output (MIMO) interference channel (IFC), where an external eavesdropper intends to wiretap one of the legitimate wireless links. By adopting the “maxmin” fairness criteria, we define the secure precoding problem as an achievable secrecy-rate maximization problem, which is inherent nonconvex and pretty hard to deal with. To tackle the inherent complexity, we recast the original nonconvex problem into a difference-of-convex (DC) programming problem through a series of equivalent transformations. Based on these endeavors, a coordinated iterative precoding algorithm is designed to solve the achievable secrecy rate maximization problem within the framework of successive convex approximation (SCA) method. The basic idea of the proposed SCA method consists in recasting the DC-programming problem into a series of convexified subproblems, where the nonconvex parts of it are linearized to their first-order Taylor expansion. Moreover, in order to ensure the convergence of the proposed iterative algorithm, a regularization method based on the proximal point idea is also employed. Numerical simulations further show that our algorithm can achieve a satisfactory performance on the premise of ensuring convergence.

Channel is important for the wireless communication system. The channel in high-speed railway is rapid time-variation and non-stationary. This papers discusses the channel characteristic in open space scenario, and defines 4 patterns. Furthermore a channel pattern recognition algorithm is proposed using k-nearest neighbor method. Simulation results show that the proposed method performs well with high accuracy and robust.

This paper studies the price-based power control strategies for non-orthogonal multiple access (NOMA) based cognitive radio networks. The primary user (PU) profits from the secondary users (SUs) by pricing the interference power made by them. Then, SUs cooperate to maximize their total revenue at the base station (BS) with successive interference cancellation (SIC) while considering their payoff to the primary user. The pricing and power control strategies between the PU and SUs are modeled as a Stackelberg game. The closed-form expression of the optimal price for the non-uniform pricing scheme is given. The computational complexity of the proposed uniform-pricing algorithm is only linear with respect to the number of SNs. Simulation results are presented to verify the effectiveness of our proposed pricing algorithm.

This paper presents our results in deep learning (DL) based single-channel blind separation (SCBS). Here, we propose a bidirectional recurrent neural network (BRNN) based separation method which can recover information bits directly from co-frequency modulated signals after end-to-end learning. Aiming at the real-time processing, a strategy of block processing is proposed, solving high error rate at the beginning and end of each block of data. Compared with the conventional PSP method, the proposed DL separation method achieves better BER performance in linear case and nonlinear distortion case with lower computational complexity. Simulation results further demonstrate the generalization ability and robustness of the proposed approach in terms of mismatching amplitude ratios.

With the rapid increasing personalized demand of C-V2X (cellular V2X) and vehicular ad hoc networks (VANET), the hybrid application of the two vehicular communications on unlicensed spectrum is becoming a trend. However, due to channel conflicts, the coexistence issue will lead to a serious drop in QoS of vehicular users. It is a challenge to allocate the wireless resource to ensure comprehensive user experience. In this paper, in order to satisfy the personalized QoS of different users while guarantee fair coexistence, we propose a conflict mitigation scheme through user association and time allocation to jointly optimize the delay and throughput, then formulate the multi-objective optimization into a mixed integer nonlinear programming (MINLP). To solve the NP-hard problem and obtain the Pareto optimal solution efficiently, we propose a PSO-based joint optimization of delay-throughput algorithm (DT-PSO). Simulation results show that our scheme outperforms existing approaches.

The most applications of wireless sensor network have stringent requirements for communication performance. To meet applications requirements, it is crucial to measure the wireless interference between nodes, which is the major factor that reduces the performance of wireless sensor networks (WSNs). However, the key problem of accurately measuring wireless interference is that the node cannot predict the neighbor node information after the network is deployed, and thus cannot establish the correspondence between the wireless interference strength and the neighbor node. To tackle this problem, this paper presents a lightweight interference measurement algorithm for WSNs. The algorithm divides the interference measurement process into three phases. The first two phases are used to gather all two-hop neighbor information by exchanging between nodes. In the third phase, each node performs interference measurements and builds the relationship of wireless interference between nodes. The experimental results show that our proposed approaches can obtain accurate inter-node wireless interference strength with low energy and communication overhead.

The structure of the network in the real world is very complex, as the dynamic network structure evolves in time dimension, how to detect network changes accurately and further locate abnormal nodes is a research hotspot. Most current feature learning methods are difficult to capture a variety of network connectivity patterns, and have a high time complexity. In order to overcome this limitation, we introduce the network embedding method into the field of network change detection, we find that node-based egonet can better reflect the connectivity patterns of the node, so a dynamic network embedding model Egonet2Vec is proposed, which is based on extracting the connectivity patterns of the node-based egonets. After the dynamic network representation learning, we use a dynamic network change detection strategy to detect network change time points and locate abnormal nodes. We apply our method to real dynamic network datasets to demonstrate the validity of this method.

In this paper, we propose a robust received signal strength (RSS) based localization method in mixed line-of-sight/non-line-of-sight (LOS/NLOS) environments, where additional path losses caused by NLOS signal propagations are included. Considering that the additional path losses vary in a dramatic range, we express the additional path losses as the sum of a balancing parameter and some error terms. By doing so, we formulate a robust weighted least squares (RWLS) problem with the source location and the balancing parameter as unknown variables, which is, simultaneously, robust to the error terms. By employing the S-Lemma, the RWLS problem is transformed into a non-convex optimization problem, which is then approximately solved by applying the semidefinite relaxation (SDR) technique. The proposed method releases the requirement of knowing specific information about the additional path losses in the previous study. Simulation results show that the proposed method works well in both dense and sparse NLOS environments.

With the development of technology, the mobile communication system has the characteristics of high rate and low delay. How to deal with the signal quickly and accurately has become a research hotspot. As the first step of the mobile communication system, the efficiency and performance of synchronization directly determine the follow-up signal Processing. In the mobile communication system, the terminal needs to synchronize the frequency and time of the received signal, that is, the synchronization signal is captured and processed. Frequency synchronization mainly carries on the digital down-conversion operation to the signal, the time synchronization is mainly through sliding the baseband signal with the locally generated synchronization sequence to determine the starting position of the synchronization signal, so as to achieve the time synchronization. Therefore, in this paper, taking LTE-A (Long Term Evolution Advanced) system as an example, a low-complexity sliding correlation method based on Fast Fourier Transform (FFT) is proposed in this paper, which can significantly reduce the computations in the synchronization process the complexity.

Physical Random Access Channel (PRACH) in 5G new radio (NR) systems transmits random access preamble for the user equipment (UE) to access the network. In 5G NR systems, Zadoff-Chu (ZC) sequences are used as random access preamble sequences. Frequency offset severely affects the perfect autocorrelation properties of the preamble sequences, thereby affecting the preamble detection performance and timing accuracy. In this paper, frequency offset effect on PRACH preamble miss detection rate and timing error in 5G NR systems is analyzed. Firstly, the frequency offset effect on inter-carrier interference and the correlation of general sequences is derived. Then, based on the former derivation and characteristics of ZC sequences, the frequency offset effect on correlation of ZC sequences is derived. Moreover, PRACH preamble miss detection rate and timing error are analyzed. The analytical results show that for different random access UEs with different PRACH preamble numbers, the random access performances are differently affected by the same frequency offset. Besides, the higher miss detection rate, the smaller timing error. The simulation results show the rationality of the analysis.

To realized an energy-efficient mode selection, we apply simultaneous wireless information and power transfer (SWIPT) to D2D communications, so that D2D and cellular users can obtain energy from receiving information, and reduce the battery energy consumption in the communication process. We leverage the theory of stochastic geometry to analyze the ergodic energy harvesting (EEH) of D2D and cellular links in reuse, dedicated, and cellular communication modes. Based on the data transmission process, we obtain the expressions of the ergodic capacity (EC) of D2D and cellular users in three D2D communication modes with power splitting (PS) architectures of SWIPT, and based on this, we derive the system energy efficiency (EE). Finally, theoretical research is demonstrated through the simulation experiments. The mode selection mechanism is performed according to the energy-efficiency. The simulations show that the system EE is improved, especially for D2D communications in reuse mode by our proposed mode selection mechanism.

Contraposing the problem of high pilot overhead and poor estimation performance for OFDM system in fast time-varying channels, a novel channel estimation method based on joint-sparse basis expansion model is proposed. In order to resist the inter-carrier interference (ICI) of OFDM system over fast time-varying channel, we introduce the OTFS (Orthogonal Time Frequency Space) technique and propose an implementation scheme of OTFS system based on time-frequency domain channel estimation. Simulation results demonstrate that the proposed OTFS system has higher reliability and better adaptability than the OFDM system in high dynamic scenarios.

In order to solve the problem that the existing algorithms in large-scale networks have high complexity in adjusting power iteratively, a load balancing mechanism based on sparse matrix prediction is proposed to achieve load balancing in C-RAN architecture. In order to minimize the correlation degree of load transfer and the balance of load transfer, the optimal sparse matrix block is obtained combined with Ncut cutting algorithm to realize dimension reduction and zero removal of the load transfer matrix. After the block, the load transfer matrix of each block is recalculated, and the load transfer matrix is used to predict the load. Finally, combined with the predicted load, the power adjustment step size is determined, and the pilot signal power of each block is adjusted in parallel to achieve load balancing. The simulation results show that the load balancing mechanism can reduce the complexity of load balancing.

Focused on the issue that the low efficiency for 5G network signaling analysis and processing, a hash topology under a new architecture based on the traditional LTE-A signaling monitoring and analysis system was proposed, its main subsystems and specific functional modules were introduced in detail, provided support for 5G terminal emulator signaling analysis. Firstly, the Key of the signaling message was sorted according to the value by using a large top heap; Secondly, the Key was mapped to a hash table, and the position of the Key value in the linked list was determined according to the probability, and the probability was obtained. The larger Key value was placed in the hash table with less conflicts. Finally, the hash table record was accessed, and the same signaling process information of the same user was associated and synthesized. The experimental results show that the improved algorithm under the proposed new architecture reduces the time spent on signaling analysis by 55.66% compared with the traditional algorithm, so it is suitable for practical engineering applications.

With the rapid development of the fifth-generation mobile communication technology (5G), existing digital signal processors (DSP) on the market cannot efficiently provide the performance required by some applications. In this situation, we design a new DSP with faster speed, lower latency and higher performance. In this article, based on the new DSP which can adapt to the new technology of 5G, we designed an assembler called Swift Assembler (SA). Different from the traditional assembler, SA is based on the Gnu Architecture Description Language, (GADL). We perform semantic analysis on GADL description files and then with the help of flex, bison and Binutils, the assembler is compiled and generated. With the support of GADL, SA has a clearer architecture and better scalability. At the same time, it covered the underlying implementation. Benefit from this, programmers can modify its source code with no need to understand the underlying implementation process. In this way, the design of interdependent hardware and software can be more easily.