Bench Testing - Solectek skyway User Manual

4. Bench testing

Before mounting units into their final location, it is recommended that the system be bench tested to
verify basic operation. The following bench test steps are suggested:
Setup. Each radio should be connected and configured per the previous Sections, with a laptop or
PC connected to each radio directly (or through a hub/switch).
Multipoint units are not factory paired like PTP kits and entries for each Client must be added to ACL
table. Be sure that Access Control MAC addresses are correct and that units share the same
bandwidth, data rate and security settings if ACL is enabled.
It is also important to have identified and prepared the antenna, RF coax and Cat5 solutions that will
be used in the intended application.
Positioning. It is important to remember that the SkyWay radio and antenna system generate and
transmit a great deal of RF power. During bench testing, antennas should not be pointed directly at
each other. Rather, establish a position so there is approximately 180 degrees angular separation
and 6 to 10 feet between units. Fine tune the antenna position so that the Local RSSI is between -
30 and -60 dBm.
Testing. If the system has been properly configured, the radios will begin communicating
immediately. The following steps are recommended to verify operation:
Link State. On the Main Status screen, verify that the RF Link State is Green (connected).
•
Local ping. From each laptop/PC be sure a ping to the local radio is successful.
•
Link ping. Now ping from one laptop/PC to the other laptop/PC. This will verify the end-to-
•
end link.
Traffic test. Using Iperf or equivalent utility, verify traffic can be passed successfully across
•
the link.
Notes:
(a) Keep in mind that the SkyWay Access data rates will stress the performance of the PC
hardware, operating system and IP stack. To ensure that this test equipment is not a performance
bottleneck, pre-testing PCs, by connecting them directly to each other, is strongly recommended.
(b) Using a file transfer to a shared volume or an FTP session on a typical Windows/Intel machine is
not adequate to accurately measure throughput.
(c) Units bench tested in an indoor configuration should not be expected to deliver full
rated throughput. Benchmarking is typically performed after a system is deployed.
9