Testing: Ubiquiti Wireless IP Video (NanoStation)

BE
Benros Emata
•Published Dec 18, 2010 05:00 AM

While the increasing maturity of IP cameras makes using IP wireless networks easier, the lack of inexpensive IP wireless transmission products is a barrier to expanding wireless IP video surveillance. Many offerings cost a $1,000 or more per link (see our test results of AvaLAN for an example).

In the midst of this, a 'cult' favorite among installers is growing from a manufacturer named Ubiqiuti. At least in the surveillance industry, Ubiquiti has zero marketing and about as low name recognition (consider them the anti-Firetide). However, they offer incredibly cheap wireless products designed for use with IP surveillance cameras.

We are always interested in learning about and sharing information on products that are less expensive and less known. To that end, we purchased a pair of Ubiquiti's NanoStation M5s to see for ourselves. We did a series of tests using up to 3 megapixel cameras simultaneously in both MJPEG and H.264 code connecting to a VMS system.

Ubiquiti Nanostation M5 is an attractive low cost solution for wireless IP video surveillance applications. Ubiqiuti can deliver multi-megapixel video transmission for about $200 per link compared to many offerings that are up to $1,000 per link.

We recommend Ubiquiti be used in less demanding commercial security projects (e.g., retail, hospitality, condo etc.) with fairly simple wireless topologies (LoS, short distances, Point to Point, and minimal bandwidth). These use cases will not require costly path-loss studies.

Higher security environments tend to demand a more reliable 'carrier class' solution, which will involve path-loss studies to determine reliability and other low-level details (azimuth, tilt, antenna selection/height, EIRP, etc.). These professional services can add considerable cost to the project.

Finally, we never were able to speak with a human being at Ubiquiti despite trying numerous times over multiple months. If support is important (and it should be for all the factors that impact wireless) and you are not absolutely confident (or brave) in going it alone, factor this in carefully for your decision to use Ubiquiti.

Key Findings

Here are key findings from our test and review of the product

  • Under 'ideal' conditions the units provided strong performance with multiple megapixel cameras streaming across link
  • We could never reach technical support by phone, only forum support seems available
  • Unit requires special adapter to accept PoE power
  • Built-in Ethernet port will not power an 802.3af IP camera
  • Unit not IP66 rated
  • Unit is UV protected for outdoor installations, but sealant should be used around cable glands
  • Highly competitive price-point (under $200 a link) to connect IP cameras to LANs wirelessly

Product Overview

The Ubiquiti Nanostation M5 (online $85) radios use two techniques to increase bandwidth:

  • Increased channel width - Uses 40MHz wide instead of 20MHz wide frequency channel
  • 2x2 MIMO (Multiple-Input/Multiple-Output) - Doubles the number of receive and transmit antennas

The Ubiquiti M5 family of radios include these 5GHz models:

  • Rocket M5 - (online $90) Provides antenna connectors for flexible topologies
  • Nanostation Loco M5 - (online $80) Compact version with slightly lower gain antenna and transmitter power
  • Bullet M5 - (online $80) A small form-factor non-MIMO for lower bandwidth applications and non-integrated antenna

The M5 series of 5GHz radios all feature the same basic chipset and AirOS operating system/interface. These radios interoperate to provide topology flexibility such as in PtP and PtMP deployments - from basic to more advanced topologies.

The Nanostation M5 interoperates with other Ubiquiti M5 radios such as the Bullet M5, Nanostation Loco M5 and Rocket M5. (All M5 radios priced similarly ~$85 online). The Bullet M5 and Rocket M5 allow for external antennas. As such they can be deployed indoors while the antenna can be mounted outdoors. The trade-off to this design is the attenuation entailed by the additional length of RF cable used between radio unit and antenna. The Nanostation M5 and Nanostation M5 Loco can be deployed outdoors, but like most integrated antenna/radio units on the market, they should be weather-sealed at the network/RF cable connectors.

Physical Overview

The Nanostation M5 radios require a proprietary PoE injector. An adapter called, Instant 802.3af [link no longer available] (MSRP $19) is available to allow the radio to accept 802.3af PoE power. The unit also provides a built-in Ethernet port so that a device can plug directly into the radio (e.g., an IP camera). The port can also be enbaled to allow PoE passthrough, but it is not 802.3af compliant and supports only a limited number of IP cameras. For outdoor deployments, the unit is UV protected (sun exposure), but cable entry area should be sealed to protect against rain/splash etc.

Administration/Configuration

The units feature the AirOS operating system, which can be accessed using a web-browser. In this video, we demonstrate the key configuration items, and some basic monitoring capabilities within the AirOS web application. We show how we configure the two radios to form a secure WPA2-PSK point-to-point link. In the next section we'll load the link with some IP cameras and examine the performance in terms of video quality and reliability.

Practical Test for IP Video

We connected the following megapixel cameras at one end of the link:

  • 1 Axis P1347 (5MP)
  • 1 Axis Q1755 (720p)
  • 1 Sony SNC-CH140 (720p)

At the other end of the link we used Exaqc's VMS to simulate a 'real-life' deployment with live viewing, and other management functionality. Also, we used the most bandwidth intensive compression type (MJPEG) to get a better sense of the load the link would be able to handle. Note that the radios are in close proximity and have LoS to each other in a well controlled lab setting.

In this video, we discover that 50+ mbps produces no detectable video quality issues. We then change from MJPEG to H.264, and observe an 80% to 90% reduction in bandwidth utilization. Again, there are no detectable performance issues or video quality degradation.