Search

In this paper, we propose the use of a prediction approach for real-time tracking in VSNs. Our solution anticipates the target’s future direction and concentrates the tracking process in specific zones to enhance the tracking precision. Furthermore, it optimizes the network load by concentrating tracking data in the predicted zones and thereby lightens the overhead on the rest of the network. To achieve this, the prediction module calculates probabilities of undertaking each direction at intersections on the basis of three parameters: road type, checkpoints avoidance and direction trends. We have conducted a stimulative study considering two types of movement profile : known and unknown profile respectively. Simulation results show a prediction success rate above 60% and up to 85% for the known profile scenarios. In addition, the performance comparison with a previous solution revealed a significant decrease in terms of network overhead, reaching a network load of 3080,622 bits/s using the prediction module as compared to 5006,88 bits for the previous protocol.

Orphan transactions are those whose parental income sources are missing at the time that they are processed. These transactions typically languish in a local buffer until they are evicted or all their parents are discovered, at which point they may be propagated further. To date, there has been little work in the literature on characterizing the nature and impact of such orphans, and yet it is intuitive that they should affect the performance of the Bitcoin network. This work thus seeks to methodically research such effects through a measurement campaign on live Bitcoin nodes. Our data show that about 45% of orphan transactions end up being included in the blockchain. Surprisingly, orphan transactions tend to have fewer parents on average than non-orphan transactions, and their missing parents have a lower fee, larger size, and lower transaction fee per byte than all other received transactions. Moreover, the network overhead incurred by these orphan transactions can be significant, exceeding 17% when using the default orphan memory pool size (i.e., 100 transactions), although this overhead can be made negligible, without significant computational or memory demands, if the pool size is simply increased to 1000 transactions. Finally, we show that when a node with an empty mempool first joins the network, 25% of the transactions that it receives become orphan, whereas in steady-state this quantity drops to about 1%.

In this work, we proposed an integration of Federated Learning with Apache Kafka, an open-source framework that enables the management of continuous data streams with fault tolerance, low latency, and horizontal scalability. Our main focus is to evaluate the impact of learning delays and network overhead when hundred of users are sending their model updates for the aggregation to improve the global model in Federated Learning.

The large-scale collection of mobile trajectories in mobile social networks makes it possible for us to use artificial intelligence, including deep learning, to explore the hidden attributes of the data and redesign data forwarding algorithms. In this paper, a data forwarding algorithm based on deep learning is proposed to transform data package communication from opportunistic forwarding to fixed path forwarding. First, by compiling statistics on real traces, we find that the number of connected nodes decreases linearly with the decrease of the sampling period, making it possible to use deep learning to process the node meeting data. Next, we design the recurrent neural network with an LSTM (Long Short-Term Memory) structure – a supervised deep learning system – to predict the probability of nodes meeting. We further propose a deep learning data forwarding algorithm which makes full use of fixed paths composed of instantaneous high-probability links. Finally, simulation results show that the algorithm proposed in this paper can effectively improve packet delivery ratio while greatly reducing network overhead.

One of the most known mechanism to ensure privacy is the use of anonymity which means that drivers have to change their identity as much as possible. In the meantime, each message has to be authenticated when it is sent. To do so, each driver should sign its messages using its private key and has to send its certificate along with the message. The use of one certificate per driver will not ensure privacy since a certificate is easily correlated to the driver identity. For this reason, each driver will use a set of certificates denoted pseudonyms certificates (PC) Each vehicle has a set of PCs which will be used to sign sent messages and will be used as a part of the driver id. A PC is used for a limited period of time. In this paper we allow to a vehicle to generate additional PCs to its neighbors when they are close together for a period of time. It will allow to these neighbors to have more PCs. The neighbors will be able to change their PCs more frequently. Our choice is to allow a pseudonym derivation when a minimum number of vehicles are driving close to each other. The disadvantage of such a method is the network overhead which will increase due to sending of the driver vehicle certificate in each sent message.

International audience; Software Defined Networking (SDN) is considered as a key paradigm for future vehicular networks and its adoption may pave the way to novel approaches to support emerging and legacy ITS services, such as Alert Message (AM) dissemination. Historically, to meet the low dissemination delay requirements and full vehicle coverage, Vehicular Ad-hoc NeTworks (VANETs) were considered as the underlying assumption of most proposed AM dissemination schemes. With SDN, vehicular-to-infrastructure (V2I) links are a viable alternative that adds to Vehicle-to-vehicle (V2V) links. Moreover, a global view of the network at a centralized controller complemented with information on vehicles and their route helps to adjust the dissemination procedure to improve its performance and efficiency. In this paper, we propose a location-based AMs dissemination scheme that combines V2I broadcasts with V2V rebroadcasts in order to provide a low delivery delay, very limited collisions, high information coverage with insignificant signaling and network overhead. The originality of our scheme stands in the selection process of V2V rebroadcasts which is based on vehicles’ location with respect to predefined rebroadcast points selected by the controller. Our proposal is evaluated and compared to legacy techniques. It shows significant improvements in delivery delays and network resource utilization.

Many of the researchers have done great effort in the field of Delay Tolerant Networks (DTNs). In this paper, all the major routing protocols (first contact, direct delivery, epidemic, prophet, max-prop and spray and wait routing protocols) are likewise investigated by varying the nodes traffic and buffer size. It is obvious from the outcomes shown that no one routing protocol is adequate for all the circumstances and diverse situation. Moreover, as the node traffic increases, Epidemic performs worst in case of Network Overhead. Additionally, as the node traffic increases, Prophet routing protocol performs better. So as to assess the performance, delivery probability ratio, network overhead ratio, avg. latency, avg. hop count and avg. buffer occupancy metrics are utilized.

In Trusted Computing, the client platform is checked for its trustworthiness using Remote Attestation. Integrity Measurement Architecture (IMA) is a well-known technique of TCG based attestation. However, due to static nature of IMA, it cannot be aware of the runtime behavior of applications which leads to integrity problems. To overcome this problem several dynamic behavior-based attestation techniques have been proposed that can measure the run-time behavior of applications by capturing all system-calls produced by them. In this paper, we have proposed a system call based technique of intrusion detection for remote attestation in which macros are used for reporting. Macros are used to denote subsequences of system calls of variable length. The basic goal of this paper is to shorten the number of system calls by the concept of macros which ultimately reduces the processing time as well as network overhead.

Orphan transactions are those whose parental income-sources are missing at the time that they are processed. These transactions are not propagated to other nodes until all of their missing parents are received, and they thus end up languishing in a local buffer until evicted or their parents are found. Although there has been little work in the literature on characterizing the nature and impact of such orphans, it is intuitive that they may affect throughput on the Bitcoin network. This work thus seeks to methodically research such effects through a measurement campaign of orphan transactions on live Bitcoin nodes. Our data show that, surprisingly, orphan transactions tend to have fewer parents on average than non-orphan transactions. Moreover, the salient features of their missing parents are a lower fee and larger size than their non-orphan counterparts, resulting in a lower transaction fee per byte. Finally, we note that the network overhead incurred by these orphan transactions can be significant, exceeding 17% when using the default orphan memory pool size (100 transactions). However, this overhead can be made negligible, without significant computational or memory demands, if the pool size is merely increased to 1000 transactions.

To improve data transmission efficiency of the apron network, multiple copies of packets are injected into the network. However, this method will result in more redundant data in the network, which further leads to network congestion. To solve the congestion, a Congestion Control Strategy based on Irregular Cellular Automaton (CCSICA) is introduced in this paper. In the apron environment, according to the interaction characteristics of the cellular automaton and surrounding neighbors, the existence of the replica of the neighboring node is fully considered. Then design the corresponding state transfer rules to achieve the purpose of more rational use of limited network resources. The simulation results show that the network overhead ratio and message delivery ratio can be improved dramatically by our proposed strategy.

The network coding optimization based on niche genetic algorithm can observably reduce the network overhead of encoding technology, however, security issues haven't been considered in the coding operation. In order to solve this problem, we propose a network coding optimization scheme for niche algorithm based on security performance (SNGA). It is on the basis of multi-target niche genetic algorithm(NGA)to construct a fitness function which with k-secure network coding mechanism, and to ensure the realization of information security and achieve the maximum transmission of the network. The simulation results show that SNGA can effectively improve the security of network coding, and ensure the running time and convergence speed of the optimal solution.

Network Function Virtualization (NFV) is emerging for service deployment in datacenters and how to efficiently deploying virtual network functions (VNFs) on VMs under SLA requirements is an key topic. Previous studies often overlook the network factor that deteriorates the datacenter performance due to unnecessary network occupation and inter-VM communication in a service function chain over the datacenter network, thus motivating us to develop Parcel-Fit. The concept of Parcel-Fit considers server capacities (i.e., CPU and memory) and network status (i.e. network occupation and server distance) to avoids unnecessary inter-VM communication in the datacenter network and thus minimize total latencies of the service function chain. In addition, Parcel-Fit is also incorporated into prior latency-minimized schemes for deploying service function chains (i.e. Multi-layer Bin Structure and SOVWin) and fulfill more SLA requirements. According to our experimental results, Parcel-Fit outperforms two conventional VM placement strategies (Best-Fit and First-Fit) on the basis of SOVWin/Multilayer Bin Structure by 28.50% less average latency, 24.54% less network occupation, 25.16% more acceptance rate and 72.04% less violation count, achieving better performance for deploying service function chains in datacenters.

Congestion is considered as a potential cause for increased transmission delay when packets are relayed between the source and destination of a mobile ad hoc environment. Thus, balancing the network resources in a distributed manner remains an indispensable consideration in facilitating effective sharing. The efficient sharing of network resources avoids load balancing in mobile nodes, thus preventing their rapid depletion. The proposed method improves the packet delivery rate and throughput of the network and it enables the process of load scheduling based on steady-state equations. Initially, it estimates the minimum queue length based on individual routes, then it sorts the node-disjoint individual routes and finally forwards the packets through the node-disjoint paths identified as reliable. This facilitates improved performances of about 21%, 17% and 25% in terms of packet delivery ratio, throughput and network overhead, compared to the benchmarked baseline queue-based load balancing techniques.

Community detection has received considerable attention because of its applications to many practical problems in mobile networks. However, when considering temporal information associated with a community (i.e., transient community), most existing community detection methods fail due to their aggregation of contact information into a single weighted or unweighted network. In this paper, we propose a contact-burst-based clustering method to detect transient communities by exploiting pairwise contact processes. In this method, we formulate each pairwise contact process as a regular appearance of contact bursts, during which most contacts between the pair of nodes happen. Based on this formulation, we detect transient communities by clustering the pairs of nodes with similar contact bursts. Since it is difficult to collect global contact information at individual nodes, we further propose a distributed method to detect transient communities. In addition to transient community detection, we also propose a new data forwarding strategy for delay tolerant networks, in which transient communities serve as the data forwarding unit. Evaluation results show that our strategy can achieve a much higher data delivery ratio than traditional community-based strategies with comparable network overhead.

— In a Wireless Sensor Network, the most significant parameter of sensor nodes is Energy. One of the demanding concerns in WSNs is Energy Efficiency. The solution to this problem is data aggregation which will decrease the network traffic. Data aggregation can be achieved by using Clustering Schemes which is an essential research area, because it provides multiple benefits such as it upgrades constancy and reduces the network overhead that prolongs the lifespan of the network. The major idea behind the clustering scheme is to deploy the sensor nodes in the network, compress the data gathered from the sensor nodes and then transfers the gathered data to the sink. Cluster head selection is the fundamental part in this procedure. Several Cluster Head Selection Techniques are considered in this paper. This paper proposed a fuzzy interface system which has used to choose a cluster head on the basis of different input parameters given to it. In this method, automated method has followed for the selection. Experimental analysis have performed between proposed, SEP and LEACH protocol on the basis of number of alive nodes, number of dead nodes and throughput. From the analysis, it has been concluded that the proposed technique enhances the lifetime of the network in comparison with other traditional methods

With the advent of cloud computing era and the dramatic increase in the amount of data applications, personalized recommendation technology is increasingly important. However, due to large scale and distributed processing architecture and other characteristics of cloud computing, the traditional recommendation techniques which are applied directly to the cloud computing environment will be faced with low recommendation precision, recommended delay, network overhead and other issues, leading to a sharp decline in performance recommendation. To solve these problems, the authors propose a personalized recommendation collaborative filtering mechanism RAC in the cloud computing environment. The first mechanism is to develop distributed score management strategy, by defining the candidate neighbors (CN) concept screening recommended greater impact on the results of the project set. And the authors build two stage index score based on distributed storage system, in order to ensure the recommended mechanism to locate the candidate neighbor. They propose collaborative filtering recommendation algorithm based on the candidate neighbor on this basis (CN-DCF). The target users are searched in candidate neighbors by the nearest neighbor k project score. And the target user's top-N recommendation sets are predicted. The results show that in the cloud computing environment RAC has a good recommendation accuracy and efficiency recommended.

We have proposed a distributed storage system which dynamically makes storage tiers and optimizes location of data blocks autonomously. This aims to enhance the I/O performance of the storage system without remarkable network overhead. Our system dynamically organizes storage tiers considering device characteristics. And the data blocks will be placed in a suitable storage tier according to their access pattern. In this paper, we propose a method to select the destination storage node for migration using an access characteristic of a data block to be migrated. This method dynamically ranks the storage nodes, and autonomously configures storage tiers based on relative performance to the migration initiator. Simulation results show that our scheme can shorten the execution time of a program with file I/O by 80% at maximum, in comparison with the static migration without considering the access characteristics of the migration data.

Recently, Named Data Networking (NDN) has been used for vehicular networks to improve the efficiency of data forwarding. However, existing solutions for vehicular NDN (V-NDN) rarely consider the priority of data packets. In this paper, we propose a priority-based data packet forwarding algorithm. We observed that different data types may have different priorities and the importance of a data packet decreases as the distance between data packet location and data source location increases. Our algorithm handles data packets differently according to their priorities. It first assigns data packets into three priority levels: high, medium and low. Then, when transmitting a data packet, the node uses either push or pull-based methods depending on the data packet's priority. For high-priority data packets, the push-based forwarding algorithm is adopted; for medium-and low-priority data packets, the pull-based forwarding algorithm is employed. With our design, the push-based forwarding algorithm is used in the local area to improve the hit ratio of high priority data; the pull-based forwarding algorithm is used to disseminate data packets to interested nodes and to reduce network overhead. Our algorithm satisfies the vehicle's need for different priority. Compared with existing works, our algorithm has a higher hit ratio than pure pull-based NDN and costs less overhead than traditional IP architectures. Simulations via SUMO, OpenStreetMap and ndnSIM verify the advantages of our work.