Learning on the Fly: An RNN-Based Online Throughput Prediction Framework for UAV Communications

This paper presents learning on the fly (LoF), a two-stage online framework to predict the achievable application-layer throughput in the downlink data communication from an unmanned aerial vehicle (UAV) to a ground access point. LoF is based on a recurrent neural network (RNN). While the UAV is flying, LoF trains the RNN with constantly observed throughput and in the meantime, makes throughput predictions for the near future. Both the training and prediction can concurrently run on a non-GPU device at the network edge (e.g., on the UAV). To this end, we design LoF with a lightweight RNN architecture and a customized training process by weighted sampling on a sliding window. We implement LoF using PyTorch. Numerical results show that LoF is able to achieve an average prediction accuracy of 87.65%, outperforming existing approaches in the literature.

[1]  Gaogang Xie,et al.  Session Throughput Prediction for Internet Videos , 2016, IEEE Communications Magazine.

[2]  Vijay K. Bhargava,et al.  An Energy-Efficient Dual Prediction Scheme Using LMS Filter and LSTM in Wireless Sensor Networks for Environment Monitoring , 2019, IEEE Internet of Things Journal.

[3]  Yan Liu,et al.  An Empirical Study of Throughput Prediction in Mobile Data Networks , 2014, 2015 IEEE Global Communications Conference (GLOBECOM).

[4]  Christian Bettstetter,et al.  Experimental analysis of multipoint-to-point UAV communications with IEEE 802.11n and 802.11ac , 2015, 2015 IEEE 26th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC).

[5]  Feng Jiang,et al.  Optimization of UAV Heading for the Ground-to-Air Uplink , 2011, IEEE Journal on Selected Areas in Communications.

[6]  Rui Zhang,et al.  Wireless communications with unmanned aerial vehicles: opportunities and challenges , 2016, IEEE Communications Magazine.

[7]  Yang Xiao,et al.  Performance analysis of priority schemes for IEEE 802.11 and IEEE 802.11e wireless LANs , 2005, IEEE Transactions on Wireless Communications.

[8]  Rui Zhang,et al.  Energy-Efficient UAV Communication With Trajectory Optimization , 2016, IEEE Transactions on Wireless Communications.

[9]  Chuck Yoo,et al.  Formula-based TCP throughput prediction with available bandwidth , 2010, IEEE Communications Letters.

[10]  Jing Wang,et al.  Spatiotemporal modeling and prediction in cellular networks: A big data enabled deep learning approach , 2017, IEEE INFOCOM 2017 - IEEE Conference on Computer Communications.

[11]  Paul Barford,et al.  A Machine Learning Approach to TCP Throughput Prediction , 2007, IEEE/ACM Transactions on Networking.

[12]  Vijay Kumar,et al.  Time-optimal UAV trajectory planning for 3D urban structure coverage , 2008, 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[13]  Paolo Dini,et al.  Mobile Traffic Prediction from Raw Data Using LSTM Networks , 2018, 2018 IEEE 29th Annual International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC).

[14]  Sofie Pollin,et al.  Micro aerial vehicle networks: an experimental analysis of challenges and opportunities , 2014, IEEE Communications Magazine.

[15]  Qi He,et al.  On the predictability of large transfer TCP throughput , 2005, SIGCOMM '05.

[16]  Keiichi Chono,et al.  Adaptive Bitrate Control of Scalable Video for Live Video Streaming on Best-Effort Network , 2018, 2018 IEEE Global Communications Conference (GLOBECOM).

[17]  Qingqing Wu,et al.  Joint Trajectory and Communication Design for Multi-UAV Enabled Wireless Networks , 2017, IEEE Transactions on Wireless Communications.

[18]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[19]  Chris C. Squires,et al.  Measurement and Characterization of Low-Altitude Air-to-Ground MIMO Channels , 2016, IEEE Transactions on Vehicular Technology.

[20]  Kenji Kanai,et al.  TRUST: A TCP Throughput Prediction Method in Mobile Networks , 2018, 2018 IEEE Global Communications Conference (GLOBECOM).

[21]  Josep Mangues-Bafalluy,et al.  Validation of the IEEE 802.11 MAC model in the ns3 simulator using the EXTREME testbed , 2010, SimuTools.

[22]  Karin Anna Hummel,et al.  Route or Carry: Motion-Driven Packet Forwarding in Micro Aerial Vehicle Networks , 2017, IEEE Transactions on Mobile Computing.

[23]  Nei Kato,et al.  A Deep-Learning-Based Radio Resource Assignment Technique for 5G Ultra Dense Networks , 2018, IEEE Network.

[24]  Ian T. Foster,et al.  Predicting the performance of wide area data transfers , 2002, Proceedings 16th International Parallel and Distributed Processing Symposium.

[25]  Hiroshi Yoshida,et al.  Constructing stochastic model of TCP throughput on basis of stationarity analysis , 2013, 2013 IEEE Global Communications Conference (GLOBECOM).

[26]  Qingqing Wu,et al.  Fundamental Trade-offs in Communication and Trajectory Design for UAV-Enabled Wireless Network , 2018, IEEE Wireless Communications.

[27]  Abbas Jamalipour,et al.  Modeling air-to-ground path loss for low altitude platforms in urban environments , 2014, 2014 IEEE Global Communications Conference.

[28]  Yi Sun,et al.  CS2P: Improving Video Bitrate Selection and Adaptation with Data-Driven Throughput Prediction , 2016, SIGCOMM.

[29]  Jianle Chen,et al.  Overview of SHVC: Scalable Extensions of the High Efficiency Video Coding Standard , 2016, IEEE Transactions on Circuits and Systems for Video Technology.

[30]  Kenji Kanai,et al.  Evaluation of Throughput Prediction for Adaptive Bitrate Control Using Trace-Based Emulation , 2019, IEEE Access.

[31]  George V. Tsoulos,et al.  LTE Ground-to-Air Field Measurements in the Context of Flying Relays , 2019, IEEE Wireless Communications.

[32]  Theodore Antonakopoulos,et al.  CSMA/CA performance under high traffic conditions: throughput and delay analysis , 2002, Comput. Commun..