TY - GEN
T1 - Spatiotemporal modeling and prediction in cellular networks
T2 - 2017 IEEE Conference on Computer Communications, INFOCOM 2017
AU - Wang, Jing
AU - Tang, Jian
AU - Xu, Zhiyuan
AU - Wang, Yanzhi
AU - Xue, Guoliang
AU - Zhang, Xing
AU - Yang, Dejun
N1 - Funding Information:
Jing Wang, Jian Tang, Zhiyuan Xu and Yanzhi Wang are with Department of Electrical Engineering and Computer Science, Syracuse University, Syracuse, NY 13244. Guoliang Xue is with Ira A. Fulton Schools of Engineering, Arizona State University, Tempe, AZ 85287. Xing Zhang is with the Key Lab of Universal Wireless Communications, Beijing University of Posts and Telecommunications (BUPT) and Beijing Advanced Innovation Center for Future Internet Technology, Beijing University of Technology (BJUT). Dejun Yang is with Department of Electrical Engineering and Computer Science, Colorado School of Mines, Golden, CO 80401. This research was supported by NSF grants 1443966, 1525920, 1461886 and 1457262. The information reported here does not reflect the position or the policy of the federal government. Xing Zhang is funded by China NSFC grants 61372114 and 61631005, and the Beijing Nova Program: Z151100000315077.
PY - 2017/10/2
Y1 - 2017/10/2
N2 - In this paper, we propose to leverage the emerging deep learning techniques for spatiotemporal modeling and prediction in cellular networks, based on big system data. First, we perform a preliminary analysis for a big dataset from China Mobile, and use traffic load as an example to show non-zero temporal autocorrelation and non-zero spatial correlation among neighboring Base Stations (BSs), which motivate us to discover both temporal and spatial dependencies in our study. Then we present a hybrid deep learning model for spatiotemporal prediction, which includes a novel autoencoder-based deep model for spatial modeling and Long Short-Term Memory units (LSTMs) for temporal modeling. The autoencoder-based model consists of a Global Stacked AutoEncoder (GSAE) and multiple Local SAEs (LSAEs), which can offer good representations for input data, reduced model size, and support for parallel and application-aware training. Moreover, we present a new algorithm for training the proposed spatial model. We conducted extensive experiments to evaluate the performance of the proposed model using the China Mobile dataset. The results show that the proposed deep model significantly improves prediction accuracy compared to two commonly used baseline methods, ARIMA and SVR. We also present some results to justify effectiveness of the autoencoder-based spatial model.
AB - In this paper, we propose to leverage the emerging deep learning techniques for spatiotemporal modeling and prediction in cellular networks, based on big system data. First, we perform a preliminary analysis for a big dataset from China Mobile, and use traffic load as an example to show non-zero temporal autocorrelation and non-zero spatial correlation among neighboring Base Stations (BSs), which motivate us to discover both temporal and spatial dependencies in our study. Then we present a hybrid deep learning model for spatiotemporal prediction, which includes a novel autoencoder-based deep model for spatial modeling and Long Short-Term Memory units (LSTMs) for temporal modeling. The autoencoder-based model consists of a Global Stacked AutoEncoder (GSAE) and multiple Local SAEs (LSAEs), which can offer good representations for input data, reduced model size, and support for parallel and application-aware training. Moreover, we present a new algorithm for training the proposed spatial model. We conducted extensive experiments to evaluate the performance of the proposed model using the China Mobile dataset. The results show that the proposed deep model significantly improves prediction accuracy compared to two commonly used baseline methods, ARIMA and SVR. We also present some results to justify effectiveness of the autoencoder-based spatial model.
KW - Autoencoder
KW - Big Data
KW - Cellular Network
KW - Deep Learning
KW - Recurrent Neural Network
KW - Spatiotemporal Modeling
UR - http://www.scopus.com/inward/record.url?scp=85034052310&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85034052310&partnerID=8YFLogxK
U2 - 10.1109/INFOCOM.2017.8057090
DO - 10.1109/INFOCOM.2017.8057090
M3 - Conference contribution
AN - SCOPUS:85034052310
T3 - Proceedings - IEEE INFOCOM
BT - INFOCOM 2017 - IEEE Conference on Computer Communications
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 1 May 2017 through 4 May 2017
ER -