How Edge Computing and Initial Congestion Window Affect Latency of Web-Based Services: Early Experiences with Baidu?

More and more things, generating huge data, will come into and enrich our lives, and the Web of Things (WoT) as a guide allows these things to be part of the World Wide Web (WWW), by using various data analysis services on the WWW. However, based on our observation on the image recognition and searching service of Baidu, pure image data transmission costs hundreds of milliseconds, besides the time of connection establishment. Inspired by the emerging Edge Computing, we analyzed the relationship between time consumption and different service provider's locations, as well as different initial congestion windows of the Transmission Control Protocol (TCP), which affect web-based services' performance. Based on our experiments in different scenarios (i.e., initial congestion window, speed of connection device and server location), we found that pushing services to the edge of network and increasing initial congestion window, both of them can reduce latency on connection establishment and data transmission, especially when users are traveling at a high speed.

[1]  Katherine Guo,et al.  Cachier: Edge-Caching for Recognition Applications , 2017, 2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS).

[2]  Quan Zhang,et al.  Firework: Data Processing and Sharing for Hybrid Cloud-Edge Analytics , 2018, IEEE Transactions on Parallel and Distributed Systems.

[3]  Paramvir Bahl,et al.  The Case for VM-Based Cloudlets in Mobile Computing , 2009, IEEE Pervasive Computing.

[4]  Raja Lavanya,et al.  Fog Computing and Its Role in the Internet of Things , 2019, Advances in Computer and Electrical Engineering.

[5]  Dawei Li,et al.  DeepCham: Collaborative Edge-Mediated Adaptive Deep Learning for Mobile Object Recognition , 2016, 2016 IEEE/ACM Symposium on Edge Computing (SEC).

[6]  V. Jacobson,et al.  Congestion avoidance and control , 1988, CCRV.

[7]  Dan Pei,et al.  TCP WISE: One initial congestion window is not enough , 2017, 2017 IEEE 36th International Performance Computing and Communications Conference (IPCCC).

[8]  Donald F. Towsley,et al.  Modeling TCP Reno performance: a simple model and its empirical validation , 2000, TNET.

[9]  Amit Agarwal,et al.  An argument for increasing TCP's initial congestion window , 2010, CCRV.

[10]  Erik Wilde,et al.  From the Internet of Things to the Web of Things: Resource-oriented Architecture and Best Practices , 2011, Architecting the Internet of Things.

[11]  Weisong Shi,et al.  Edge Computing: Vision and Challenges , 2016, IEEE Internet of Things Journal.

[12]  Aakanksha Chowdhery,et al.  Networked Drone Cameras for Sports Streaming , 2017, 2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS).

[13]  Katherine Guo,et al.  Precog: prefetching for image recognition applications at the edge , 2017, SEC.

[14]  Hong Zhong,et al.  Demo Abstract: EVAPS: Edge Video Analysis for Public Safety , 2016, 2016 IEEE/ACM Symposium on Edge Computing (SEC).