Multi-step network traffic prediction using echo state network with a selective error compensation strategy

With the increasing complexity of the network, the current network traffic has strong nonlinearity and burstiness. Therefore, the traditional traffic prediction model is no longer applicable. The neural network model, especially the LSTM, can well fit the nonlinearity of time-series data and preserve the information memory of the past. However, as for the periodicity of long-term span network traffic data, the neural network model does not perform well. Based on this, this paper proposes LTS-TP (Long-Term Span Traffic Prediction model), a network traffic prediction model, to solve the problem. First, the model decomposes the collected network traffic data using the improved STL decomposition algorithm to preserve the seasonal component. Then, the trend component and the remainder component are input into the Seq2Seq model based on the LSTM added with the improved attention mechanism for prediction. Finally, the predicted value of the output is added to the seasonal component, and the final network traffic prediction value is obtained. In the simulation part, this paper uses the MAWI public data set to test the proposed network traffic prediction model and compared performance with other models. The results show that the network traffic prediction model proposed in this paper has a good predictive effect on long-term span network traffic data.

Echo state network (ESN) has recently attracted increasing interests because of its superior capability in modeling nonlinear dynamic systems. In the conventional echo state network model, its dynamic reservoir (DR) has a random and sparse topology, which is far from the real biological neural networks from both structural and functional perspectives. We hereby propose three novel types of echo state networks with new dynamic reservoir topologies based on complex network theory, i.e., with a small-world topology, a scale-free topology, and a mixture of small-world and scale-free topologies, respectively. We then analyze the relationship between the dynamic reservoir structure and its prediction capability. We utilize two commonly used time series to evaluate the prediction performance of the three proposed echo state networks and compare them to the conventional model. We also use independent and identically distributed time series to analyze the short-term memory and prediction precision of these echo state networks. Furthermore, we study the ratio of scale-free topology and the small-world topology in the mixed-topology network, and examine its influence on the performance of the echo state networks. Our simulation results show that the proposed echo state network models have better prediction capabilities, a wider spectral radius, but retain almost the same short-term memory capacity as compared to the conventional echo state network model. We also find that the smaller the ratio of the scale-free topology over the small-world topology, the better the memory capacities.

Traffic flow prediction is critical for Intelligent Transportation Systems to alleviate congestion and optimize traffic management. The existing basic Encoder-Decoder Transformer model for multi-step prediction requires high computational complexity, making them less efficient for real-time applications. Therefore, this paper presents an Encoder-Only Transformer that eliminates the decoder component while evaluating the model’s ability to capture long-term spatial-temporal dependencies. This simplification was found to reduce computational overhead while maintaining a reasonable predictive accuracy comparable to the basic Encoder-Decoder Transformer and Long Short-Term Memory (LSTM) models. Experimental evaluations on two real-world datasets, Minnesota and California, for hourly prediction of 6, 12, and 24-hour horizon tasks, demonstrate the proposed model’s effectiveness. The proposed Encoder-Only Transformer outperforms LSTM networks across all horizons task for the Minnesota dataset, achieving up to 17.33% improvement in Mean Absolute Error. The proposed model excels in 24-hour horizon task for the California dataset but underperforms for shorter horizons compared to the LSTM and basic Encoder-Decoder Transformer models. These findings highlight the Encoder-Only Transformer’s application for multi-step traffic flow prediction while emphasizing dataset-specific variations in model performance.

Substorms are a magnetospheric phenomenon which causes high geomagnetic and auroral disturbances. It is widely accepted that substorm activity is controlled by solar wind conditions. It is, however, difficult to predict substorms deterministically because of the complex physical processes underlying substorm occurrences. We propose a framework for modelling time series of event occurrences controlled by external forcing. In this framework, occurrences of external-driven events are represented with a non-stationary Poisson process, and its intensity, which corresponds to the occurrence rates per unit time, is described with a simple machine learning model, the echo state network, which is fed with forcing variables. The echo state network is trained by maxmising the likelihood given the event time series data. We apply this approach for analysing time series of substorm onsets identified from Pi2 pulsations, which are irregular geomagnetic oscillations associated with substorm onsets. We train the echo state network to well describe the response of substorm activity to solar-wind conditions. We then examine the characteristics of the substorm activity by feeding synthetic solar-wind data into the echo state network. The results show what solar wind variables effectively contribute to the substorm occurrence. Our echo state network model is also useful for examining the statistical properties of the substorm occurrence rate. For example, we can evaluate what mainly controls the seasonal and UT variations of substorm activity. There are two explanations for the seasonal and UT variations. One explanation is that the seasonal and UT variations is controlled by the inner product between the solar-wind magnetic field and the Earth's dipole axis. The other is that the variations are due to the angle between the solar-wind flow and the Earth's dipole axis. Since these two effects are related with different input variables in our echo state network model, we can examine the contribution of each effect to the substorm occurrence frequency. The result shows that the seasonal and UT variations are mostly dependent on the angle between the solar-wind flow and the Earth's dipole axis. 

Internet of Things (IoT) is a network of interconnected devices, such as sensors and smart devices that have processing, sensing, and communication capabilities, as well as can transmit information to each other and a supreme console through the Internet. Network traffic prediction is an important operational and management function for any data network. It has a significant role in today's increasingly complex and diverse networks. Network traffic prediction is also more important for IoT networks to provide reliable communication. The artificial neural network (ANN) has been successfully applied to traffic prediction. In this paper, we perform the IoT traffic time series prediction using a multistep ahead prediction with Time Series NARX Feedback Neural Networks. The estimation error of a prediction approach has been evaluated using the performance functions MSE, SSE, and MAE, besides, another measure of prediction accuracy the mean absolute percent of error (MAPE).

Can connectome-derived constraints inform computation? In this paper we investigate the contribution of a fruit fly connectome's topology on the performance of an Echo State Network (ESN) - a subset of Reservoir Computing which is competitive in chaotic time series prediction. Specifically, we replace the reservoir layer of a classical ESN - normally a fixed, random graph represented as a 2-d matrix - with a particular (female) fruit fly connectome-derived connectivity matrix. We refer to this experimental class of models (with connectome-derived reservoirs) as "Fruit Fly ESNs" (FFESNs). We train and validate the FFESN on a chaotic time series prediction task; here we consider four sets of trials with different training input sizes (small, large) and train-validate splits (two variants). We compare the validation performance (Mean-Squared Error) of all of the best FFESN models to a class of control model ESNs (simply referred to as "ESNs"). Overall, for all four sets of trials we find that the FFESN either significantly outperforms (and has lower variance than) the ESN; or simply has lower variance than the ESN.

SummaryIn order to improve the prediction accuracy of network traffic, a novel prediction model based on ensemble empirical mode decomposition and multiple models is proposed. In this study, ensemble empirical mode decomposition algorithm is introduced to decompose the network traffic and get several components. Approximate entropy is introduced to judge the complexity of each component. According to the results of approximate entropy, echo state network is selected to predict high complexity components, support vector machine is used to predict medium complexity components, and autoregressive integrated moving average model is introduced to predict low complexity components. The advantages of each prediction model are used to predict the appropriate component. The final prediction results are obtained by adding the predicted values of each component. In order to solve the problem that the prediction performance of support vector machine and echo state network is affected by their parameters, an improved whale optimization algorithm is proposed to optimize the parameters of the model. Meanwhile, the calculation results of approximate entropy show that compared with the original network traffic, the complexity of each component obtained by ensemble empirical mode decomposition is reduced, which reduces the complexity of modeling. Three network traffic datasets with sampling periods of 10 ms, 1 s, and 10 min are collected. Compared with the other five state‐of‐the‐art prediction models, the case study results show that the proposed prediction model has better prediction accuracy and excellent statistical performance indicators.

As the fundamental part of other Intelligent Transportation Systems (ITS) applications, short-term traffic volume prediction plays an important role in various intelligent transportation tasks, such as traffic management, traffic signal control and route planning. Although Neural-network-based traffic prediction methods can produce good results, most of the models can’t be explained in an intuitive way. In this paper, we not only proposed a model that increase the short-term prediction accuracy of the traffic volume, but also improved the interpretability of the model by analyzing the internal attention score learnt by the model. we propose a spatiotemporal attention mechanism-based multistep traffic volume prediction model (SAMM). Inside the model, an LSTM-based Encoder-Decoder network with a hybrid attention mechanism is introduced, which consists of spatial attention and temporal attention. In the first level, the local and global spatial attention mechanisms considering the micro traffic evolution and macro pattern similarity, respectively, are applied to capture and amplify the features from the highly correlated entrance stations. In the second level, a temporal attention mechanism is employed to amplify the features from the time steps captured as contributing more to the future exit volume. Considering the time-dependent characteristics and the continuity of the recent evolutionary traffic volume trend, the timestamp features and historical exit volume series of target stations are included as the external inputs. An experiment is conducted using data from the highway toll collection system of Guangdong Province, China. By extracting and analyzing the weights of the spatial and temporal attention layers, the contributions of the intermediate parameters are revealed and explained with knowledge acquired by historical statistics. The results show that the proposed model outperforms the state-of-the-art model by 29.51% in terms of MSE, 13.93% in terms of MAE, and 5.69% in terms of MAPE. The effectiveness of the Encoder-Decoder framework and the attention mechanism are also verified.

Network traffic prediction performs a main function in characterizing network community performance. An approach which could appropriately seize the salient characteristics of the network visitors could be very useful for network analysis and simulation. Network traffic prediction methods could be divided into two classes: one is the single models and the opposite is the hybrid fashions. The hybrid models integrate the merits of several single models and consequently can enhance the network traffic prediction accuracy. In this paper, a new hybrid network traffic prediction method (EPSVM) primarily based on Empirical Mode Decomposition (EMD), Particle Swarm Optimization (PSO), and Support Vector Machines (SVM) is presented. The EPSVM first utilizes EMD to eliminate the impact of noise signals. Then, SVM is applied to model training and fitting, and the parameters of SVM are optimized by PSO. The effectiveness of the presented method is examined by evaluating it with different methods, including basic SVM (BSVM), Empirical Mode Decomposition processed by SVM (ESVM), and SVM optimized by Particle Swarm Optimization (PSVM). Case studies have demonstrated that EPSVM performed better than the other three network traffic prediction models.

In order to reveal the internal dynamic property of short-term traffic flow, the nonlinear analysis method is used to identify the chaotic property of traffic flow which is the basis for the prediction of the traffic flow time series. Traffic flow time series is reconstructed in phase-space based on chaos theory. The embedding dimension and delay time are first calculated via the C-C method. The correlative dimension of attractor is then calculated with the Grassberger-Procaccia method. The largest Lyapunov exponent of traffic flow set is calculated on the basis of the improved small data set method to verify the presence of the chaos in traffic flow time series. A novel multi-step adaptive prediction method is proposed to solve the problem of adjusting the filter parameters of the chaos local adaptive prediction method during traffic flow multi-step prediction. The traffic flow time series are found to have chaotic properties in different statistical scales of 2, 4, and 5 min and show that the improved small data set method can accurately evaluate the chaotic property for traffic flow time series, and that the multi-step adaptive prediction method is capable of effectively predicting its fluctuation, which provides a useful reference for traffic guidance and control.

In this paper, a comparison of two network traffic activities prediction models will be presented, namely a backpropagation neural network (BPNN) and a genetic algorithm based backpropagation neural network (GABPNN). A backpropagation neural network (BPNN) prediction model can be used to learn a time-series dataset. However, the performance of the BPNN can be improved by optimizing the BPNN using a genetic algorithm. This paper outlines a network traffic prediction model that is developed using a typical backpropagation neural network (BPNN) coupled with a genetic algorithm (GA). The performance of the GABPNN is measured by using a statistical analysis, namely Mean of Square Error (MSE). The results show that the MSE value of GABPNN obtained was lower with the population size of 400, crossover probability was 0.01, uniform mutation probability was 0.8, and 50 iterations. Therefore, the obtained results of the GABPNN model illustrates that the proposed model GABPNN has improved the prediction accuracy compared to the traditional BPNN model.

The real-time intelligent perception and prediction of traffic situation can assist connected automated vehicles (CAVs) in path planning and reduce traffic congestion in Cognitive Internet of Vehicles (CIoVs). The centralized traffic congestion prediction solutions generally fail to adapt to the dynamic traffic environment and lead to significant communication overheads. Blockchain technology has attracted great attention in the information sharing of vehicular networks for its advantages in decentralization, transparency, traceability, and tamper-proof capability. However, due to the bottlenecks, such as high computational cost, current blockchains are incapable actuate on efficient online traffic situational cognition and prediction for CIoVs. Motivated by this, we propose a blockchain-enabled cognitive segments sharing framework for online multistep congestion duration prediction. We design a cognitive model of traffic situation based on anomaly detection and filtering mechanism to guarantee the accuracy of the cognitive segments before being packaged into the block. Furthermore, to improve the consensus efficiency, we design a credit evaluation mechanism and propose a credit-based delegated Byzantine fault tolerance (CDBFT) algorithm. Finally, we propose an online multistep prediction algorithm based on long short-term memory (LSTM) to predict future traffic congestion duration. Experimental results demonstrate that the proposed algorithms achieve shorter consensus latency and higher predictive accuracy than the existing algorithms.

Accurate prediction of network traffic is crucial to network resource management, whether it's network resource provisioning or malicious attack detection, reliable network traffic prediction mechanisms are needed for advance warning existing network traffic prediction methods still have some obvious limitations, existing network traffic prediction methods still have some obvious limitations, such as high computational complexity, VMD decomposition parameter optimization problem, premature convergence into local optimum solutions, and integrated prediction model prediction accuracy is not high many problems. In response to the above, this paper presents a model that combines Variation Mode Decomposition (VMD) and Extreme Learning Machine (ELM) for predicting network traffic based on Genetic Algorithm (GA).

With the growing popularity of smart phones and the rapid development of Internet, the traditional network management system cannot adapt itself to the requirement intelligently. If we can use historic data to predict the trend of network traffic accurately, a better planning of network is more likely to be made and limited resource can be allocated and scheduled reasonably. However, massive amounts of data collected by network operators cannot be effectively processed. Therefore, in this paper, we design a network traffic prediction system based on Hadoop platform to process the real mobile network traffic data for a major network operator in China. With the Echo State Network (ESN), a new kind of Recurrent Neural Network (RNN) structure, the system can make accurate predictions of traffic variation trend for various network applications.

Traffic prediction is of significant importance for telecommunication network planning and network optimization. In this paper, the traffic series from a certain mobile network in Heilongjiang province in China is studied. The characteristics in respect of both trend and periodicity are explored with autocorrelation function. Based on the characteristics exhibited in the traffic series, multiplicative seasonal autoregressive integrated moving average model (ARIMA) is employed to make traffic series prediction. Average daily traffic per month for the province as well as its every sub-region from July to December in 2009 is forecasted and compared with the actual operation data. The mean absolute percentage error (MAPE) for one-step ahead prediction is 1.382%, and MAPE for the 6 steps is within 6%. The prediction result is of high precision. Furthermore, the cause for the big prediction error in 2 regions is analyzed, and the appropriateness of the model is testified on the opposite aspect. This paper also provides an effective method by using correlation coefficients to analyze the cause for significant prediction errors which do not happen in time series prediction applications rarely.