Technology

A test for mobile broadband speed measurement

Description
1. A test for  mobile broadband speed  measurement Toshiya Jitsuzumi Kyushu University Jitsuzumi@econ.kyushu‐u.ac.jp Feb. 12, 2014 2. Test Setting (1/2) ã…
Categories
Published
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  • 1. A test for  mobile broadband speed  measurement Toshiya Jitsuzumi Kyushu University Jitsuzumi@econ.kyushu‐u.ac.jp Feb. 12, 2014
  • 2. Test Setting (1/2) • Installed an original measurement application on 11 iPhones  and measured QoS on a 10 minute interval for 2 hours. • Date and time Jan. 24, 2014 16:00‐18:00 (Japan Time) Jan. 31, 2014 16:00‐18:00 (Japan Time) • Measurement  A classroom at the 5th floor of the building of  the Faculty of Economics at Kyushu University. Each iPhone cut Wi‐Fi connection and used only  LTE. The measurement server was located at Tokyo.
  • 3. Test Setting (2/2) • Our application conducted QoS measurements every 10  minutes for 2 hours. • 12 measurement windows for each handset. • In each measurement window • RTT was measured 150 times at the interval of 0.2 seconds. • Download/Upload speed was calculated based on throughputs for  10 seconds. • Upload speed was measured once. • Download speeds were measured in three different settings:  one on a single TCP stream, the others are on 3 and 5 TCP  streams. • The number of hops was counted 3 times using “traceroute.”
  • 4. Result 1: Averages Date Median RTT (second) N Upload Speed (Mbps) Download Speed (Mbps) Number of hops 1 TCP stream 3 TCP streams 5 TCP streams 0.0782 2.02 5.78 5.69 5.87 14.62 13 0.1085 0.82 1.91 1.85 1.42 14.08 iPhone5 16 0.0536 3.01 8.92 8.82 9.49 15.06 SoftBank average 95 0.0582 9.84 16.50 18.64 18.81 24.84 iPhone5 69 0.0556 11.01 16.75 18.36 19.10 25.97 iPhone5S Jan. 31 29 iPhone4S Jan. 24 26 0.0653 6.72 15.85 19.37 18.02 21.85 17 0.0700 4.92 10.66 11.14 11.10 14.55 iPhone4S 5 0.0970 8.23 8.98 11.60 11.31 13.33 iPhone5 12 0.0588 3.54 11.36 10.95 11.02 15.06 SoftBank average 82 0.0596 15.03 17.32 17.86 18.69 20.82 iPhone5 61 0.0597 14.52 17.90 18.11 20.00 23.62 iPhone5S 21 0.0593 16.50 15.66 17.14 14.87 12.68 KDDI average KDDI average
  • 5. Result 2: Standard Deviations Date Median RTT (second) N Upload Speed (Mbps) Download Speed (Mbps) Number of hops 1 TCP stream 3 TCP streams 5 TCP streams 0.0280 1.64 3.93 4.09 4.61 0.66 13 0.0038 0.08 0.80 0.52 0.53 0.58 iPhone5 16 0.0075 1.64 2.35 2.89 3.02 0.19 SoftBank average 95 0.0133 5.79 8.59 9.85 10.57 5.14 iPhone5 69 0.0119 5.46 7.49 7.01 7.99 3.49 iPhone5S Jan. 31 29 iPhone4S Jan. 24 26 0.0142 5.45 10.96 14.93 15.44 7.21 17 0.0258 7.80 8.53 8.61 8.64 1.07 iPhone4S 5 0.0242 13.63 14.49 14.81 14.93 1.29 iPhone5 12 0.0166 1.51 3.72 3.67 3.58 0.18 SoftBank average 82 0.0134 12.97 10.36 11.65 10.70 7.29 iPhone5 61 0.0136 8.99 7.94 9.98 9.12 6.32 iPhone5S 21 0.0127 20.47 15.24 15.49 13.64 1.43 KDDI average KDDI average
  • 6. Result 4: Coefficient of Variations Date Median RTT (second) N Upload Speed (Mbps) Download Speed (Mbps) Number of hops 1 TCP stream 3 TCP streams 5 TCP streams 0.36 0.81 0.68 0.72 0.79 0.05 13 0.04 0.10 0.42 0.28 0.37 0.04 iPhone5 16 0.14 0.55 0.26 0.33 0.32 0.01 SoftBank average 95 0.23 0.59 0.52 0.53 0.56 0.21 iPhone5 69 0.21 0.50 0.45 0.38 0.42 0.13 iPhone5S Jan. 31 29 iPhone4S Jan. 24 26 0.22 0.81 0.69 0.77 0.86 0.33 17 0.37 1.58 0.80 0.77 0.78 0.07 iPhone4S 5 0.25 1.66 1.61 1.28 1.32 0.10 iPhone5 12 0.28 0.43 0.33 0.34 0.32 0.01 SoftBank average 82 0.22 0.86 0.60 0.65 0.57 0.35 iPhone5 61 0.23 0.62 0.44 0.55 0.46 0.27 iPhone5S 21 0.21 1.24 0.97 0.90 0.92 0.11 KDDI average KDDI average
  • 7. Result 5: Median RTT (second) 0.14 16:00‐, Jan. 24, 2014 0.14 16:00‐, Jan. 31, 2014 0.12 0.12 0.10 0.10 0.08 0.08 KDDI‐4S 0.06 KDDI‐5 0.06 SB‐5 KDDI‐5 0.04 SB‐5S 0.04 SB‐5S 0.02 0.02 0.00 0.00 950 1000 1050 TIME(s) 1100 1150 1950 KDDI‐4S SB‐5 2000 2050 TIME(s) 2100 2150
  • 8. Result 6: Upload Speed (Mbps) 20.0 16:00‐, Jan. 24, 2014 20.0 18.0 18.0 16.0 16.0 14.0 14.0 12.0 8.0 12.0 KDDI‐4S 10.0 KDDI‐5 8.0 SB‐5 6.0 SB‐5S 6.0 4.0 4.0 2.0 2.0 0.0 0.0 1950 16:00‐, Jan. 31, 2014 10.0 950 1000 1050 TIME(s) 1100 1150 KDDI‐4S KDDI‐5 SB‐5 SB‐5S 2000 2050 TIME(s) 2100 2150
  • 9. Result 7: Download Speed on 1 TCP stream(Mbps) 50.0 16:00‐, Jan. 24, 2014 50.0 40.0 40.0 30.0 16:00‐, Jan. 31, 2014 30.0 KDDI‐4S KDDI‐4S KDDI‐5 20.0 SB‐5 SB‐5S 20.0 10.0 KDDI‐5 SB‐5S SB‐5 10.0 0.0 950 1000 1050 TIME(s) 1100 1150 0.0 1950 2000 2050 TIME(s) 2100 2150
  • 10. Result 8: Download Speed on 3 TCP streams(Mbps) 50.0 16:00‐, Jan. 24, 2014 50.0 40.0 40.0 30.0 16:00‐, Jan. 31, 2014 30.0 KDDI‐4S KDDI‐4S KDDI‐5 20.0 SB‐5 SB‐5S 20.0 10.0 KDDI‐5 SB‐5S SB‐5 10.0 0.0 950 1000 1050 TIME(s) 1100 1150 0.0 1950 2000 2050 TIME(s) 2100 2150
  • 11. Result 8: Download Speed on 5 TCP streams(Mbps) 50.0 16:00‐, Jan. 24, 2014 50.0 40.0 40.0 30.0 16:00‐, Jan. 31, 2014 30.0 KDDI‐4S KDDI‐4S KDDI‐5 20.0 SB‐5 SB‐5S 20.0 10.0 KDDI‐5 SB‐5S SB‐5 10.0 0.0 950 1000 1050 TIME(s) 1100 1150 0.0 1950 2000 2050 TIME(s) 2100 2150
  • 12. Findings • Softbank’s LTE network offered faster but more volatile speed than  KDDI’s LTE. • Since the variations in the measurement outputs were significant, a  small number of speed measurements could not produce a  meaningful outcome. • Throughputs changed quite dramatically in a short amount of time. • Considering the size of standard deviation, the difference that comes  from the number of TCP streams in the download speed  measurement was not statistically significant, and therefore can be  ignored in a practical sense.
  • sdm pro

    Sep 19, 2017
    Search
    Related Search
    We Need Your Support
    Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

    Thanks to everyone for your continued support.

    No, Thanks