On the importance of TCP splitting proxies for future 5G mmWave communications

5G mmWave technology promises capacities 10 to 100 times that of 4G. However, mmWave links are very sensitive to having direct line of sight between sender and receiver, with dramatically fluctuating capacities due to transient blocking and shadow fading, which can substantially degrade TCP performance. TCP-splitting proxies (SPEPs) are often used to improve TCP performance over wireless links. We investigate current operator SPEPs (used in 4G LTE networks) under mmWave-like dynamics using the MONROE1 testbed. We introduce a mmWave emulation model for an urban canyon scenario, and use it to evaluate how current operational SPEPs impact the performance. Our results from experiments including four operators and three countries, indicate that SPEPs will be even more important for 5G mmWave than they are in LTE. Current 4G SPEPs provide significant, though not optimal, benefits under mmWave-like conditions, allowing them to be utilized as the network transitions to 5G technology.

[1]  Sundeep Rangan,et al.  End-to-End Simulation of 5G mmWave Networks , 2017, IEEE Communications Surveys & Tutorials.

[2]  Korian Edeline,et al.  Hic Sunt Proxies: Unveiling Proxy Phenomena in Mobile Networks , 2019, 2019 Network Traffic Measurement and Analysis Conference (TMA).

[3]  Ramesh Govindan,et al.  Investigating Transparent Web Proxies in Cellular Networks , 2015, PAM.

[4]  Seung-Woo Ko,et al.  Enhancing TCP end-to-end performance in millimeter-wave communications , 2017, 2017 IEEE 28th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC).

[5]  Joerg Widmer,et al.  Analysis of TCP Performance in 5G mm-Wave Mobile Networks , 2019, ICC 2019 - 2019 IEEE International Conference on Communications (ICC).

[6]  Theodore S. Rappaport,et al.  Rapid Fading Due to Human Blockage in Pedestrian Crowds at 5G Millimeter-Wave Frequencies , 2017, GLOBECOM 2017 - 2017 IEEE Global Communications Conference.

[7]  Olav N. Østerbø.  Scheduling and capacity estimation in LTE , 2011, International Teletraffic Congress.

[8]  Andreas Kassler,et al.  TCP performance over 5G mmWave links — Tradeoff between capacity and latency , 2017, 2017 IEEE 13th International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob).

[9]  Marco Ajmone Marsan,et al.  Experience: An Open Platform for Experimentation with Commercial Mobile Broadband Networks , 2017, MobiCom.

[10]  Gabriel Montenegro,et al.  Performance Enhancing Proxies Intended to Mitigate Link-Related Degradations , 2001, RFC.

[11]  Viktor Farkas,et al.  A Split Connection TCP Proxy in LTE Networks , 2012, EUNICE.

[12]  Marco Ajmone Marsan,et al.  Speedtest-Like Measurements in 3G/4G Networks: The MONROE Experience , 2017, 2017 29th International Teletraffic Congress (ITC 29).

[13]  Seung-Woo Ko,et al.  Exploiting Caching for Millimeter-Wave TCP Networks: Gain Analysis and Practical Design , 2018, IEEE Access.

[14]  Karen Kafadar,et al.  Letter-Value Plots: Boxplots for Large Data , 2017 .

[15]  Ilyas Alper Karatepe,et al.  Big data caching for networking: moving from cloud to edge , 2016, IEEE Communications Magazine.

[16]  Van Jacobson,et al.  Controlling queue delay , 2012, Commun. ACM.

[17]  Jing Zhu,et al.  milliProxy: A TCP proxy architecture for 5G mmWave cellular systems , 2017, 2017 51st Asilomar Conference on Signals, Systems, and Computers.

[18]  Amr Rizk,et al.  A measurement study on the application-level performance of LTE , 2014, 2014 IFIP Networking Conference.

[19]  Alessio Botta,et al.  Monitoring and measuring wireless network performance in the presence of middleboxes , 2011, 2011 Eighth International Conference on Wireless On-Demand Network Systems and Services.

[20]  Jing Zhu,et al.  Will TCP Work in mmWave 5G Cellular Networks? , 2018, IEEE Communications Magazine.