|
|
Asking the Right Questions
When Considering an 802.11n Test Solution
Whether you are developing an IEEE 802.11n Access Point (AP), or about to make a major purchasing decision and need to evaluate offerings from various network equipment vendors, having the right tools to analyze the performance of the network is critical.
Here are some questions that need to be answered when considering
test equipment to evaluate complex 802.11n based wireless local area networks:
Does the equipment provide a complete test package?
A complete test package consists of the following components:
Traffic Generation
The questions that need to be answered when evaluating a traffic generator are:
- Does the solution offer traffic generation capabilities?
- If not, do you have to rely on an external source for generation?
- What will I use for my external traffic generation source?
- Can an external traffic generation source load the system I am testing to full air-rate?
- Is an external traffic generation source that supports 802.11n even available? Is it compatible with my System Under Test (SUT)? How much does that add to the overall cost? How will it keep up with the 802.11n standards?
- How will I create repeatable scenarios for functional testing? What application will I use to control the external traffic source?
- If I have problems with my external traffic generation source and its compatibility with my tester and/or my SUT, who will help me resolve them?
- How can you test the behavior of the SUT with multiple clients associated and passing traffic if the test environment only supports one individual client?
Traffic Analysis
The questions that need to be answered when evaluating a traffic analyzer are:
- Is a comprehensive analyzer included? One that includes counters for every layer 2,3 and 4 frame type, as well as for errored frames?
- If the solution lacks counters, what other means exist to report test results?
- Is payload integrity checking provided by the test solution? Otherwise, how will the test equipment report data corruption by the SUT? Will there be a need for yet another external device to analyze traffic passing through SUT and to the wired (Ethernet) side of the network?
- Does the test solution provide analysis of packet reordering? Otherwise, how will the tester detect potential frame re-ordering or frame duplication by the SUT (especially when encountering buffer boundary conditions)?
- Is precise latency measurement provided? The requirement that the SUT offers a minimal latency (e.g. less than 50mS) is critical in voice and video applications.
- Does the tester offer frame loss measurement – another key component for support of triple-play applications. Are those offered on a per flow basis?
Traffic Capture
The questions that need to be answered when considering traffic capture capabilities are:
- How many spatial streams are supported by the traffic capture device? A MIMO test solution has to support four (4) spatial streams to accommodate 4x4 situations.
- Is the solution based upon an off-the-shelf chipset? Is so, how does that affect its traffic capture abilities?
- Will it be able to capture bad and/or corrupted packets? If not, how will it be possible to determine that these are present and how will it be possible to debug the SUT?
- What version of the IEEE 802.11n standard does the chipset support? If it’s a draft (e.g. Draft 2.0), when and how will it be upgrades to the final version? How much will the upgrade cost? Will it even be compatible with the SUT?
- Does the capture capability support 802.11n as well as legacy 802.11a/b/g?
- Will the capture function handle “air-rate” for 802.11n up to MIMO 4x4? If not, how does that affect the testing as the SUT evolves to MIMO 4x4 support?
- How can RSSI levels be verified on captured frames? Does the capture mechanism record RSSI for each frame?
Channel Emulation
The questions that need to be answered when considering the channel emulation capabilities of the test system are:
- Does the test solution offer channel emulation? Can a MIMO SUT really be tested (in a cabled RF environment) without channel emulation?
- If an integrated channel emulator is not provided, how much does an external channel emulator add to the cost of the test solution? What if the SUT consists of more than one 802.11n device?
- Can the channel emulator create complete channel emulation to the IEEE six defined models or just “limited cross-path channel environment”?
- If an external channel emulator is to be used, how can the settings of this channel emulator be correlated with the test case scenarios?
- If an external channel emulator is to be used, how can the test bed be automated?
Roaming and Mobility
The questions that need to be answered when considering the ability of the test solution to verify roaming and mobility capabilities of the SUT are:
- Roaming is a client/station decision per 802.11 standards. In a testing environment that relies upon external clients that are not provided by the test equipment manufacturer, what is being tested? Is it the behavior of the client or the SUT?
- Repeatability is crucial to measuring roaming delays. How can this be guaranteed with an external client (station) that is not controlled by the test system?
- Does the test solution offer precise metrics for measuring the roam time?
- If roaming performance needs to be tested with more than one client (station), as is the case in the real world, how is this achieved if the test relies on the use of external clients? How will the test scale to dozens of clients roaming at different intervals?
Support and Commitment
Purchasing an 802.11n test solution is an expensive proposition and needs to provide a long term Return on Investment.
Beyond features and capabilities of test gear there is an equally important element of support and commitment from the test equipment vendor for a successful project delivery.
Conclusion - VeriWave offers a complete package for 802.11a/b/g and now 802.11n testing
download PDF ›
top 
|
|
|
