Cisco Cisco 1800 2800 3800 Series AP-AG 802.11a b g High-Speed WIC
© 2015 Cisco and/or its affiliates. All rights reserved. This document is Cisco/Verizon Public Information.
Page 5 of 26
LTE QoS Planning Guidelines
-
The ISR prerequisite for PNTM on Verizon 4G LTE MPN is an ISR with embedded Verizon LTE interface
running recommended IOS and modem firmware releases. All other Verizon and Cisco recommendations
for MPN and DMNR are also applicable. The following guide provides the specifics.
www.cisco.com/c/dam/en/us/td/docs/routers/access/interfaces/software/deployment/guide/guide_c07-720264.pdf
-
The physical deployment prerequisite is a site survey, which can be done before or after installation.
Antenna placement and type can make a significant difference in LTE throughput. ISRs provide details on
LTE signal strength and quality. The details change within a minute of moving the antennas. The values
can be seen from the ISR console, or by text message from a cell phone using this no charge application
https://supportforums.cisco.com/document/12316801/commands-over-sms
documented here
www.cisco.com/c/dam/en/us/td/docs/routers/access/interfaces/software/deployment/guide/guide_100413.pdf
-
The overall LTE shaped rate T (first level policy) defined on LTE egress from the ISR should be set to a
value that is always achievable. If the shaped rate is above the actual achievable rate, QoS may not be
engaged, and traffic requiring QoS (e.g. Voice over IP) may experience degradation (e.g. voice clipping,
echo, silence). Determining the overall LTE shaped rate requires a repeatable process at each location,
including peak days, times and conditions Examples of methods to do this include:
o
For a location where LTE MPN is not yet deployed, bring an ISR (e.g. 819) and laptop running a
test tool (e.g. JPERF), perform a site survey (optionally using the SMS app above), and run the
test tool with a UDP traffic stream (may require a test tool at the main location to “catch” the
stream, e.g. a PC also running JPERF with “server” radio button selected). With iterations at
higher speeds, determine at what Mbps rate the JPERF catcher sees less than what the JPERF
laptop behind the LTE ISR is sending. Also, a “show interface cellular” command (with
appropriate interface number) can confirm the value seen by the ISR. Note that doing the latter,
requires the JPERF test stream to run twice as long as the ISR LTE interface’s “load interval”
(which can be set down to 30 seconds with the ISR interface configuration command “load-
interval 30”).
o
For a location where LTE MPN is already deployed, confirm that a site survey has been
completed. If LTE is the backup path, determine how to run the test across LTE. Run a test tool
from that location with a UDP traffic stream (may require a test tool at the main location to “catch”
the stream). With iterations at higher speeds, determine at what Mbps rate the JPERF catcher
sees less than what the JPERF laptop behind the LTE ISR is sending. The ISR “show interface
cellular” command (with appropriate interface number) will be required, as the LTE interface will
account for real traffic also traversing LTE . The “load interval” caveat above is still pertinent.
Note that if there is production traffic over LTE, the test tool speeds should be set to not unduly
impact applications.
o
If the overall shaped rate T chosen is greater than the ISR’s achievable upstream LTE bit rate,
packets may be dropped without QoS being engaged. The behavior and impact on users and
applications would be the same as if no QoS was defined, e.g. best effort treatment for all traffic.