Deep in the Bowels of LTE Network Performance

Deep in the Bowels of LTE Network Performance

A BENCHMARK STUDY OF LTE NETWORK PERFORMANCE FOR THE FOUR LARGEST US MOBILE OPERATORS AND ALL OF THEIR INFRASTRUCTURE SUPPLIERS IN TWO MAJOR MARKETS

04/03/2014 | 73 pages
Price: $2,150.00
Purchase this Issue

LTE Performance involving 5 US operators and their Infrastructure Suppliers

In our continuing “Mother of all Network Benchmark Tests” series of reports, we provide results from LTE network testing that we did in San Francisco and Phoenix, Arizona, including the ASU campus in Tempe.  Operators include AT&T, Sprint, T-Mobile and Verizon Wireless.  Infrastructure vendors include Alcatel Lucent, Ericsson, Nokia Solutions and Networks, and Samsung.
This study was done in collaboration with Sanjole who provided us with its WaveJudge 4900A LTE Analyzer and IntelliJudge test platform.  With these powerful tools we were able to capture and analyze all Physical Layer and other protocol exchanges that occurred over-the-air (OTA) between an eNB and the active mobile devices that were within its coverage area.

KEY QUESTIONS THAT WE ANSWER:

  • What are the “real” user downlink / uplink data rates on today’s LTE networks, based on actual user generated traffic and what, if any, impact does network loading have on the typical user experience?
  • How loaded are today’s LTE networks and how does loading vary across networks, not to mention geographically within a single operator’s network?
  • How frequently is MIMO used and what is the typical distribution of mobile devices that use transmit diversity versus MIMO?  How and why are these metrics noticeably different between networks?
  • What impact does an operator’s network design criteria and in-building mobile devices accessing the outdoor macro network have on network performance?
  • What is the actual downlink / uplink spectral efficiency of LTE and why do factors such as user / application behavior, device performance, network design, and the distribution of mobile devices throughout the cell result in the spectral efficiency being markedly different from widely-held views?
  • What are some of the individual vendor performance traits that exist and how do they impact network performance?