Study on Characteristics of Metric-aware Multipath Algorithms in Real Heterogeneous Networks
Author | Zhang, Xiaoya |
Author | Dong, Ping |
Author | Du, Xiaojiang |
Author | Zhang, Yuyang |
Author | Zhang, Hongke |
Author | Guizan, Mohsen |
Available date | 2024-07-24T10:13:52Z |
Publication Date | 2021 |
Publication Name | 2021 IEEE Global Communications Conference, GLOBECOM 2021 - Proceedings |
Resource | Scopus |
Abstract | Multipath transmission is considered one of the promising solutions to improve wireless resource utilization where there are many kinds of heterogeneous networks around. Most scheduling algorithms rely on real-time network metrics, including delay, packet loss, and arrival rates, and achieve satisfying results in simulation or wired environments. However, the implicit premise of a scheduling algorithm may conflict with the characteristics of real heterogeneous wireless networks, which has been ignored before. This paper analyzes the real network metrics of three Chinese heterogeneous wireless networks under different transmission rates. To make the results more convincing, we conduct experiments in various scenarios, including different locations, different times of the day, different numbers of users, and different motion speeds. Further, we verify the suitability of a typical delay-aware multipath scheduling algorithm, Lowest Round Trip Time, in heterogeneous networks based on the actual data measured above. Finally, we conclude the characteristics of heterogeneous wireless networks, which need to be considered in a well-designed multipath scheduling algorithm. |
Sponsor | ACKNOWLEDGMENT This work was supported in part by the Fundamental Research Funds for the Central Universities 2020YJS019, and NSFC under grant No.61872029, No.61972026. |
Language | en |
Publisher | Institute of Electrical and Electronics Engineers Inc. |
Subject | Heterogeneous Wireless Network Measurement Metrics-aware Multipath |
Type | Conference |
Files in this item
Files | Size | Format | View |
---|---|---|---|
There are no files associated with this item. |
This item appears in the following Collection(s)
-
Computer Science & Engineering [2402 items ]