Performance Evaluation Of Networks Research Articles

Migrating From Legacy to Software Defined Networks: A Network Reliability Perspective

Designing survivable communication networks to achieve carrier-grade five-nines reliability is of paramount importance for the network operators. This article addresses service reliability and its related aspects such as nodal reachability, network connectivity, and edge-disjoint routing in both traditional networks and software defined networks (SDNs). The proposed roadmap is based on two phases: Fundamental analytical phase and performance evaluation phase. In the first phase, a graph operator is defined to analyze the characteristics of the reliability metric and its associated reachability feature. This phase will focus on both the macro- and micro-level properties of reliability. In the second phase, we exploit the analysis in the former phase to get an insight into the performance evaluation of traditional and SDN-based networks against the reliability metric, and then calculate the statistical significance of the mean difference of their reliability values. Reliability under edge-disjoint paths to avoid resource competition is also investigated. Various types of topologies are utilized to test the service reliability of both architecture designs. Extensive simulation results show that SDN-based networks have comparable performance to its legacy counterpart against the operational reliability metric. Our findings not only shed light on enhancing reliability using edge-disjoint paths under link failure scenarios but also expected to benefit the operators to achieve their service level objectives while migrating from legacy to SDN-based platform.

Flexible Pavement Crack’s Severity Identification and Classification using Deep Convolution Neural Network

Effective road maintenance program is vital to ensure traffic safety, serviceability, and prolong the life span of the road. Maintenance will be carried out on pavements when signs of degradation begin to appear and delays may also lead to increased maintenance costs in the future, when more severe changes may be required. In Malaysia, manual visual observation is practiced in the inspection of distressed pavements. Nonetheless, this method of inspection is ineffective as it is more laborious, time consuming and poses safety hazard. This study focuses in utilizing an Artificial Intelligence (AI) method to automatically classify pavement crack severity. Field data collection was conducted to allow meaningful verification of accuracy and reliability of the crack’s severity prediction based on AI. Several important phases are required in research methodology processes including data collection, image labelling, image resizing, image enhancement, deep convolution neural network (DCNN) training and performance evaluation. Throughout the analysis of image processing results, the image output was successfully classified using MATLAB software. The good agreement between field measurement data and DCNN prediction of crack’s severity proved the reliability of the system. In conclusion, the established method can classify the crack’s severity based on the JKR guideline of visual assessment.

Performance evaluation of convolutional neural network on Tianhe-3 prototype

Exascale supercomputers will greatly support the expanding computational resource demand of convolutional neural networks (CNNs). At present, the prototype cluster of Tianhe-3 supercomputer, which is based on the Chinese-made many-core processors, the Phytium-2000+ (FTP) and Matrix-2000+ (MTP), has gone into service. We evaluated the training performance of CNN on the Tianhe-3 prototype. The performance of image convolution and matrix multiplication on the FTP and MTP was tested to evaluate the single-node performance, and the Allreduce element was tested to evaluate the scalability of the distributed training on the prototype cluster. We also qualitatively analyzed the performance bottlenecks of CNN on the FTP and MTP processors by Roofline model and provided some optimization suggestions for improving the CNN on the Tianhe-3 prototype.

Performance and Scalability Evaluation of the Wireless Body Area Network using Castalia Simulator

Castalia is a model for WLAN (WSN), body areas (BAN), and low-power networks, which are commonly frequently used in research and technology, for networking devices. This paper, provides a simple assessment of Castalia’s memory use and computing time over a number of scenarios and characteristics of Wireless Body Area Networks (WBANs). While, WBAN gives you the ability to monitor patients closely, but it presents many new challenges too. HKernel parameters like network size, simulation time, and mobile node fraction are, however, unique to demonstrate the scalability potential of Castalia. We analyze our results and clarify the success of the simulator against intuitive results. Also examined in traffic adaptive approaches are IEEE 802.15.4 standard called Zigbee and IEEE 802.15.6 standard called baseline MAC. In fact, the procedures are also compared, the distribution ratio (PDR) are compared and the energy demand is evaluated. They are evaluated. Castalia users should use the tests and their interpretations as a guide to assess the drawback of their simulations, as well as to select criteria that balance performance consistency. We also demonstrate future users the efficiency capabilities of Castalia-3.2

Performance evaluation for vehicular ad-hoc networks based routing protocols

VANET is a branch of MANETS, where each vehicle is a node, and a wireless router will run. The vehicles are similar to each other will interact with a wide range of nodes or vehicles and establish a network. VANETs provide us with the infrastructure to build new solutions for improving safety and comfort for drivers and passengers. There are several routing protocols proposed and evaluated for improving VANET's performance. The simulator is preferred over external experience because it is easy, simple, and inexpensive. In this paper, we choose AODV protocol, DSDV protocol, and DSR protocol with five different nodes density. For each protocol, as regards specific parameters like (throughput, packet delivery ratio, and end- to- end delay). On simulators that allow users to build real-time navigation models of simulations using VANET. Tools (SUMO, MOVE, and NS-2) were used for this paper, then graphs were plotted for evaluation using Trace-graph. The results showed the DSR is much higher than AODV and DSDV, In terms of throughput. While DSDV is the best choice because of the low average end to end delay. From the above, we conclude that each strategy has its own negative and positive aspects that make it ideally suited to a particular scenario than other scenarios.

Performance Evaluation of GSM and WCDMA Networks: A Case Study of the University of Ilorin

Performance Evaluation of GSM and WCDMA Networks: A Case Study of the University of Ilorin

Neural network combining X-ray and ultrasound in breast examination

With the acceleration of people’s life rhythm, the incidence of breast cancer is gradually increasing. This study mainly explores the application of neural network combined with X-ray and ultrasound in breast examination. The process discussed in this research is mainly based on AdaBoost integrated neural network. According to the basic principles of artificial neural networks, first use training samples to train the neural network, then form an integrated neural network based on the AdaBoost algorithm, and then use the test samples to perform prediction tests on the network, and perform calculations in the integrated neural network to predict the classification the results are compared with the actual results in the test set, and the total number of misclassifications is analyzed to determine whether the prediction effect of the integrated neural network is good or bad. In the process of training the neural network, the analysis of the obtained case data needs to be numerically processed and then normalized to map the data to the [− 1,1] interval. The segmentation goal is to segment each tissue in the breast ultrasound image, which are skin layer, fat layer, gland layer, tumor, muscle layer, and background layer. The mean squared error is used as the network performance evaluation function. The research results are measured by the error rate. The lower the error rate, the better the classification prediction effect and the better the performance of the individual neural network. The diagnostic coincidence rate combined with X-ray and ultrasound was 88.33% (265/300). Compared with pure ultrasound or mammography, the difference was significant (P < 0.05). The recall rate of neural network combined with X-ray and ultrasound reached 91.4%. The results show that the neural network combined with X-ray and ultrasound has extremely high application value in breast examination.

Modeling and Performance Analysis of Satellite Network Moving Target Defense System with Petri Nets

With the development of satellite communication networks and the increase of satellite services, security problems have gradually become some of the most concerning issues. Researchers have made great efforts, including conventional safety methods such as secure transmission, anti-jamming, secure access, and especially the new generation of active defense technology represented by MTD. However, few scholars have theoretically studied the influence of active defense technique on the performance of satellite networks. Formal modeling and performance analysis have not been given sufficient attention. In this paper, we focus on the performance evaluation of satellite network moving target defense system. Firstly, two Stochastic Petri Nets (SPN) models are constructed to analyze the performance of satellite network in traditional and active defense states, respectively. Secondly, the steady-state probability of each marking in SPN models is obtained by using the isomorphism relation between SPN and Markov Chains (MC), and further key performance indicators such as average time delay, throughput, and the utilization of bandwidth are reasoned theoretically. Finally, the proposed two SPN models are simulated based on the PIPE platform. In addition, the effect of parameters on the selected performance indexes is analyzed by varying the values of different parameters. The simulation results prove the correctness of the theoretical reasoning and draw the key factors affecting the performance of satellite network, which can provide an important theoretical basis for the design and performance optimization of the satellite network moving target defense system.

Performance evaluation of joint relay selection and network coding for multi-source and multi-destination relay network

Network coding and relay selection have attracted significant attention due to their performance improvement and spectral efficiency in wireless networks. In this work, we study joint relay selection and network coding cooperation schemes for M source-destination pairs assisted with N relays. We propose a new smart decode-and-forward max-min relay selection scheme based on the source-destination links’ outage status and L selected relays to enhance the communication system diversity. Our objective is realize joint network coding and relay selection strategies for maximizing the expected number of recovered packets by the intended destinations in various situations. We analyze the performance of the proposed scheme and prove that the maximum diversity order of $$N+1$$ is achieved when the number of active relays is at least $$\lceil {\frac{N+1}{3}}\rceil $$ . Also, we show that the scheme can achieve a diversity order of 3L when $$L<\lceil {\frac{N+1}{3}}\rceil $$ . Our analytical and simulation results indicate that the proposed scheme provides higher performance than the existing network coded cooperation schemes.

Performance Evaluation of Aerospace Service Network Based on Entropy Weight Set-Pair-Analysis

The Aerospace Service Network (ASN) provides communication of voices, video, data. Its performance is the key to satellite launching. In order to comprehensive assessment of Aerospace Service Network’s performance, this paper presented evaluation index based on network analysis data and calculated index weight by entropy method. Furthermore, a network performance evaluation model based on interval number Set-Pair-Analysis is established. The analysis and comparison showed that this model is objective and effective, and it provides decision support for network maintenance.

Performance monitoring and evaluation of FTTx networks for 5G backhauling

One of the main challenges in today’s access networks for 5G base stations (BSs) is the final links (or the last mile) of the network. The links that constitute the main backbone of the network are mostly based on Internet Protocol (IP)/Multi-protocol Label Switching networks with connection over fiber cables. However on the access network side, Mobile Network Operators (MNOs) may also benefit from existing copper lines in addition to fiber and microwave lines. Therefore, developing a 5G deployment strategy to connect BSs that are installed indoor and outdoor locations with these different connection types to the backhaul network, will be important to ensure cost-effective installations. Utilizing Fiber-to-the-x (FTTx) networks that are already installed for broadband connectivity will provide MNOs significant initial investment benefits. In this paper, we have analyzed 5G readiness of the backhaul links of a real infrastructure provider using the proposed evaluation methodology. Our large-scale experimental analysis identifies the problematic sites (where possible solutions and suggestions are also proposed) and determines the suitability of the copper circuits for services that can be provided over the next generation networks. At the end of the paper, we also discuss about the challenges and advantages of utilizing different 5G services when FTTx is used as a backhaul and propose an ordering of the measurable FTTx specific quality parameters according to next generation service deployment scenarios.

Fuzzy Logic Approach for Routing in Internet of Things Network

A performance of network is evaluated by considering different parameters. The network lifetime depends on many factors Residual energy, Link lifetime and Delay. The Major Challenge in IoT is to the increased lifetime of low power and lossy network (RPL).The process considering input and output to evaluate Network performance by considering the above factors. The proposed system makes use of FIS (Fuzzy Inference System) for selecting the best path to maximize network lifetime. The outcome obtained by using MATLAB and Network performance is increased. The excellent route is selected if Residual Energy is 194, Link quality is 51.2 and Delay is 1.05 then excellent route quality is 73.4%.

Performance evaluation of deep neural networks for forecasting time‐series with multiple structural breaks and high volatility

The problem of automatic and accurate forecasting of time-series data has always been an interesting challenge for the machine learning and forecasting community. A majority of the real-world time-series problems have non-stationary characteristics that make the understanding of trend and seasonality difficult. Our interest in this paper is to study the applicability of the popular deep neural networks (DNN) as function approximators for non-stationary TSF. We evaluate the following DNN models: Multi-layer Perceptron (MLP), Convolutional Neural Network (CNN), and RNN with Long-Short Term Memory (LSTM-RNN) and RNN with Gated-Recurrent Unit (GRU-RNN). These DNN methods have been evaluated over 10 popular Indian financial stocks data. Further, the performance evaluation of these DNNs has been carried out in multiple independent runs for two settings of forecasting: (1) single-step forecasting, and (2) multi-step forecasting. These DNN methods show convincing performance for single-step forecasting (one-day ahead forecast). For the multi-step forecasting (multiple days ahead forecast), we have evaluated the methods for different forecast periods. The performance of these methods demonstrates that long forecast periods have an adverse effect on performance.

Performance monitoring and evaluation of FTTx networks for 5G backhauling

Performance monitoring and evaluation of FTTx networks for 5G backhauling

Network Performance and Management-Complexity Evaluation of Software-Defined Access Network in Multi-Tenancy Scenario

Industrial revolution has transformed social life. Because of the digital disruption, companies reorganize and change their business model to face the challenges of industrial revolution 4.0. The multi-tenant architecture separates each tenant logically while they are physically integrated. This model has a problem in the management of network devices. The difficulty of changing and rigidity of the underlying infrastructure presents few possibilities for innovation or improvement since network devices have generally been vendor locked. A Software-Defined Access Network (SDAN) is a software-defined network that is implemented on an access network. SDAN aims to simplify complicated access management and improve service provision. This paper aims to analyze and evaluate SDAN design performance and complexity management of network devices. Evaluation is carried out by conducting a comparative survey between SDAN and non-SDAN in terms of performance and use case scenarios. Based on the results, it can be concluded that the SDAN has 10%-20% better performance in terms of elapsed times and throughput. It also significantly reduces time for managing network devices ten times faster than non-SDAN.

Joint Statistical and Machine Learning Approach for Practical Data-Driven Assessment of User Throughput Quality in Microcellular Radio Networks

The provision of enhanced user experience with higher-data transmission speed over the radio frequency channels was one of the core objectives of evolving Long Term Evolution (LTE), which is a 4th generation of cellular broadband communication standard. LTE was developed to supports data transmission speed up to 300 Mbps and 75 Mbps in downlink and uplink, respectively, utilizing a robust multiplexing technique. For mobile radio networks operators to maintain high operability around the ever increasing and demanding subscribers, periodic evaluation, quantitative estimation and analysis of network performance is pivotal. There exist a quantum of previous research works on LTE system network performance that have been conducted both at spatial domain and temporal domain, but the authors concentrated their studies either on practical LTE radio coverage issues only or on general LTE network performance issues using analytical/simulation techniques. In this contribution, a combined statistical and machine learning approach is proposed and applied to provide an in-depth user data throughput performance of operational LTE networks in relation to signal coverage and signal quality parameters, using in typical microcellular built-up terrains as case study. Specifically, the impact of transmitter- receiver communication distance, received signal reference power, received Signal reference quality, Signal to Noise Ratio, Received Signal Strength and Channel Quality Indicator on user throughput over Radio Link Control, Physical Downlink Shared Channel and Packet Data Convergence Protocol layers have been shown using professional TEMS investigation tools. Particularly, for the purpose of effective LTE system network planning and management, the most influencing signal coverage and signal quality parameters on user throughput performance are also examined and shown in the framework of Self-organsing map (SOM) and statistical correlation stratagems.

Performance evaluation of ultra-dense macro-femto cellular networks based on efficient channel resource allocation system

Femtocells have been discovered as the solution to coverage and capacity challenges experienced by mobile users in indoor environments, where data and voice services are mostly employed. During communication, path losses are mostly experienced because of base station distances and walls separating the communicating devices, which ultimately affects coverage and quality of service adversely. This study proposes an efficient channel allocation scheme (ECA) and compared with light channel (LCA), stern channel (SCA) and open channel allocation (OCA) schemes to ascertain the most optimal. These schemes assign frequencies such that interference is diminished significantly after femtocells, femto-users, and macro-users have been deployed randomly within the macro-cell coverage area. Results show higher capacities for macro-femto edge users (105 Mbps) compared to macro-femto center users (11 Mbps) with the proposed algorithm (ECA) emerging as the most optimal in both cases. Deployment of 30 macro-femto edge users records a failure probability of 0.11 for the proposed scheme, 0.17 for OCA, 0.19 for LCA, and 0.29 for SCA which is much lower compared to center users. It was observed that 50% of the edge users recorded downlink interference of 0.8 dBm for proposed, 1.3 dBm for OCA, 1.6 dBm for LCA, and 1.7 dBm for SCA which is much reduced compared to center users.

Bayesian sample size determination in a single-server deterministic queueing system

Although queueing models in the mathematical field of queueing theory are mainly studied in the steady-state regime and practical applications are interested mostly in queues in transient situations subject to burst arrivals and congestion, their use is still justified as a first step towards a more complex and thorough analysis. In these practical applications, parameters such as the traffic intensity, which is the ratio between the arrival rate and service rate, are unknown and need to be estimated statistically. In this study, a Markovian arrival and deterministic single-server queueing system, known in Kendall notation as an M∕D∕1 queueing model, is considered. This is one of the simplest queueing models with deterministic service time and may be seen as an approximation of a variety of applications in the performance evaluation of production management, telecommunications networks, and other areas. The main goal of this manuscript is to propose a methodology to determine the sample size for an M∕D∕1 queueing system under the Bayesian setup by observing the number of customer arrivals during the service time of a customer. To verify the efficiency and efficacy of the proposed approach, an extensive set of numerical results is presented and discussed.

Performance evaluation of artificial neural networks for natural terrain classification

Remotely sensed image segmentation and classification form a very important part of remote sensing which involves geo-data processing and analysis. Artificial neural networks (ANNs) are powerful machine learning approaches that have been successfully implemented in numerous fields of study. There exist many kinds of neural networks and there is no single efficient approach for resolving all geospatial problems. Therefore, this research aims at investigating and evaluating the efficiency of three ANN approaches, namely, backpropagation neural network (BPNN), radial basis function neural network (RBFNN), and Elman backpropagation recurrent neural network (EBPRNN) using multi-spectral satellite images for terrain feature classification. Additionally, there has been close to no application of EBPRNN in modeling multi-spectral satellite images even though they also contain patterns. The efficiency of the three tested approaches is presented using the kappa coefficient, user’s accuracy, producer’s accuracy, overall accuracy, classification error, and computational simulation time. The study demonstrated that all the three ANN models achieved the aim of pattern identification, segmentation, and classification. This paper also discusses the observations of increasing sample sizes as inputs in the various ANN models. It was concluded that RBFNN’s computational time increases with increasing sample size and consequently increasing the number of hidden neurons; BPNN on overall attained the highest accuracy compared to the other models; EBPRNN’s accuracy increases with increasing sample size, hence a promising and perhaps an alternative choice to BPNN and RBFNN if very large datasets are involved. Based on the performance metrics used in this study, BPNN is the best model out of the three evaluated ANN models.

Efficiency Parameters for Node Indexing to Get Energy Efficient and Bandwidth Aware Path

Background: MANET is a constrained environment having limited energy, restricted bandwidth, and unstable wireless connectivity, etc. It creates a high demand to design an efficient QoS routing protocol having a multi-criteria metric to make efficient routing decisions. Aims: This work is to enhance end-to-end network throughput. To achieve our aim we optimized MBQA-OLSR, which is an enhancement of MP-OLSR. MP-OLSR is enhancement of OLSR, which is a prominent proactive routing protocol. Main advantageous concept in OLSR is selection of MPR nodes that minimize excessive transmission of control packets. Selected MPR nodes should carry enough battery to increase network life time. Packet loss is another issue on wirelesThe main aim of QoS based routing protocol EB-OLSR (Energy efficient and Bandwidth aware Optimized Link State Routing) is to achieve enhanced aggregate network throughput. Which is acquired by selecting QoS qualified efficient nodes to transmit data packets. EB-OLSR set out analogous efficient mobile nodes for routing by applying newly defined multi-criteria metric EPNI (Efficiency Parameters for Node Indexing) which estimates the node’s efficiency index on the basis of four QoS parameters that are unconsumed power, slothful period, queue length and residual bandwidth.s links, it is due to poor quality unstable links; hence link quality is another work of this paper. Method: We obtained a single node index for each mobile node through EPNI metric based on residual battery, node idle time, packet queue length and residual bandwidth. Result: EB-OLSR produced aggregate network throughput enhancement of 574 bits per second in comparison of MP-OLSR and 390.66 bits per second in comparison to MBQA-OLSR. Average End-to-End Delay for EB-OLSR is 0.41, 0.16 times less and Packet Delivery Ratio enhances 30.74%, 9.5% respectively. Conclusion: This paper evaluated network performance parameters of proposeExperimental results proved that the EB-OLSR is an energy-efficient and bandwidth aware optimized link-state routing protocol, which successfully meets the main aim of this study i.e. throughput enhancement of mobile ad-hoc network. The EPNI metric is the major strength of the work which is used to get a single-valued weighted sum of multiple criteria.d routing protocol EB-OLSR that are packet delivery ratio average end to end delay and throughput. It has been observed from simulation experiments that EB-OLSR outer perform other compared similar kind of routing protocols at different network sizes. In proposed work, EPNI metric is defined to get weighted sum of multiple criteria and it is considered to make decision for routing data packets. Single valued multiple criteria evaluation helps to simplify relationship and priority between multiple criteria and combined them in a single node index. Hence along with optimized throughput control overheads are also minimized. EB-OLSR is energy efficient and bandwidth aware optimized link state routing protocol, which successfully meet the main aim of this study i.e. throughput enhancement of mobile ad-hoc network. Aggregate throughput enhances 574 bits per second in comparison of MP-OLSR and 390.66 bits per second in comparison of MBQA-OLSR. Average End-to-End Delay for EB-OLSR is 0.41, 0.16 times less in comparison of MP-OLSR, MBQA-OLSR respectively. Packet delivery ratio enhances 30.74% in comparison of MP-OLSR and 9.5% in comparison of MBQA-OLSR. Performance analyzes at varying traffic load shows only slighter improvement. In this study, we put fixed weights to every considered metric according to general priority. In future, we will study the priorities of metrics in different network scenarios like at varying node speeds, at different mobility patterns, etc to see whether priorities of metrics changes in different scenario or not.