Datasets:

yuningm

/

citesum

like 7

Abstract-Short-term traffic forecasting based on deep learning methods, especially long short-term memory (LSTM) neural networks, has received much attention in recent years. However, the potential of deep learning methods in traffic forecasting has not yet fully been exploited in terms of the depth of the model architecture, the spatial scale of the prediction area, and the predictive power of spatial-temporal data. In this paper, a deep stacked bidirectional and unidirectional LSTM (SBU-LSTM) neural network architecture is proposed, which considers both forward and backward dependencies in time series data, to predict network-wide traffic speed. A bidirectional LSTM (BDLSM) layer is exploited to capture spatial features and bidirectional temporal dependencies from historical data. To the best of our knowledge, this is the first time that BDLSTMs have been applied as building blocks for a deep architecture model to measure the backward dependency of traffic data for prediction. The proposed model can handle missing values in input data by using a masking mechanism. Further, this scalable model can predict traffic speed for both freeway and complex urban traffic networks. Comparisons with other classical and state-of-the-art models indicate that the proposed SBU-LSTM neural network achieves superior prediction performance for the whole traffic network in both accuracy and robustness. Index Terms-Deep learning, bidirectional LSTM, backward dependency, traffic prediction, network-wide traffic HE performances of intelligent transportation systems (ITS) applications largely rely on the quality of traffic information. Recently, with the significant increases in both the total traffic volume and the data they generate, opportunities and challenges exist in transportation management and research in terms of how to efficiently and accurately understand and exploit the essential information underneath these massive datasets. Short-term traffic forecasting based on data driven models for ITS applications has been one of the biggest developing research areas in utilizing massive traffic data, and has great influence on the overall performance of a variety of In the last three decades, a large number of methods have been proposed for traffic forecasting in terms of predicting speed, volume, density and travel time. Studies in this area normally focus on the methodology components, aiming at developing different models to improve prediction accuracy, efficiency, or robustness. Previous literature indicates that the existing models can be roughly divided into two categories, i.e. classical statistical methods and computational intelligence (CI) approaches [2] . Most statistical methods for traffic forecasting were proposed at an earlier stage when traffic condition were less complex and transportation datasets were relatively small in size. Later on, with the rapid development in traffic sensing technologies and computational power, as well as traffic data volume, the majority of more recent work focuses on CI approaches for traffic forecasting. With the ability to deal with high dimensional data and the capability of capturing non-linear relationship, CI approaches tend to outperform the statistical methods, such as autoregressive integrated moving average (ARIMA) [36] , with respect to handling complex traffic forecasting problems [38] . However, the full potential of artificial intelligence was not exploited until the rise of neural networks (NN) based methods. Ever since the precursory study of utilizing NN into the traffic prediction problem was proposed [39], many NN-based methods, like feed forward NN [41], fuzzy NN [40], recurrent NN (RNN) [42] , and hybrid NN [25] , are adopted for traffic forecasting problems. Recurrent Neural Networks (RNNs) model sequence data by maintaining a chain-like structure and internal memory with loops [4] and, due to the dynamic nature of transportation, are especially suitable to capture the temporal evolution of traffic status. However, the chain-like structure and the depth of the loops make RNNs difficult to train because of the vanishing or blowing up gradient problems during the backpropagating process. There have been a number of attempts to overcome the difficulty of training RNNs over the years. These difficulties were successfully addressed by the Long ShortTerm Memory networks (LSTMs) [3] , which is a type of RNN with gated structure to learn long-term dependencies of Ruimin Ke is with the

GSTNet further investigated capturing global dynamic dependencies to improve in traffic network prediction tasks REF .

24760423

Deep Bidirectional and Unidirectional LSTM Recurrent Neural Network for Network-wide Traffic Speed Prediction

Wireless mesh networks are expected to be widely used to provide Internet access in the near future. In order to fulfill the expectations, these networks should provide high throughput simultaneously to many users. Recent research has indicated that, due to its conservative CSMA/CA channel access scheme and RTS/CTS mechanism, 802.11 is not suitable to achieve this goal. In this paper, we investigate throughput improvements achievable by replacing CSMA/CA with an STDMA scheme where transmissions are scheduled according to the physical interference model. To this end, we present a computationally efficient heuristic for computing a feasible schedule under the physical interference model and we prove, under uniform random node distribution, an approximation factor for the length of this schedule relative to the shortest schedule possible with physical interference. This represents the first known polynomial-time algorithm for this problem with a proven approximation factor. We also evaluate the throughput and execution time of this algorithm on representative wireless mesh network scenarios through packet-level simulations. The results show that throughput with STDMA and physical-interferencebased scheduling can be up to three times higher than 802.11 for the parameter values simulated. The results also show that our scheduling algorithm can schedule networks with 2000 nodes in about 2.5 minutes.

In a similar problem setting, Brar et al. REF presented a computationally efficient, centralized heuristic for computing a feasible schedule under the physical SINR model.

6818545

Computationally efficient scheduling with the physical interference model for throughput improvement in wireless mesh networks

Underground forums, where participants exchange information on abusive tactics and engage in the sale of illegal goods and services, are a form of online social network (OSN). However, unlike traditional OSNs such as Facebook, in underground forums the pattern of communications does not simply encode pre-existing social relationships, but instead captures the dynamic trust relationships forged between mutually distrustful parties. In this paper, we empirically characterize six different underground forumsBlackHatWorld, Carders, HackSector, HackE1ite, Freehack, and L33tCrew -examining the properties of the social networks formed within, the content of the goods and services being exchanged, and lastly, how individuals gain and lose trust in this setting.

Analysis of such underground services was first documented in REF where the authors examined the properties of social networks formed for blackmarket services.

207191376

An analysis of underground forums

Access is the killer app" 3 is the vision of the Daedalus project at U.C. Berkeley. Being able to be connected seamlessly anytime anywhere to the best network still remains an unful lled goal. Often, even determining the best" network is a c hallenging task because of the widespread deployment o f o verlapping wireless networks. In this paper, we describe a policy-enabled hando system that allows users to express policies on what is the best" wireless system at any moment, and make tradeo s among network characteristics and dynamics such a s cost, performance and power consumption. We designed a performance reporting scheme estimating current network conditions, which serves as input to the policy speci cation. A primary goal of this work is to make it possible to balance the bandwidth load across networks with comparable performance. To avoid the problem of hando instability, i.e., many mobile hosts making the same hando decision at essentially the same time, we designed randomization into our mechanism. Given the current best" network, our system determines whether the hando is worthwhile based on the hando overhead and potential network usage duration.

Wang et al. REF describe a policy based handover control system that allows users to express a policy on what is the gives a brief introduction to previous work related to mobility and handover management.

42916380

Policy-enabled handoffs across heterogeneous wireless networks