Features

As wireless signals travel through the medium, they undergo a loss in the signal strength due to signal attenuation caused by the medium and obstacles. This attenuation is exponential and the signal drops to several orders within a short distance. With the loss of signal strength, bit error rate (BER) of the transmitted data increases. Radio frequency (RF) interference from other sources can further increase the BER. This can result in sluggish performance, as the sending station may not receive acknowledgments for its transmitted data from the receiving station.

Moreover, this may require the sending station to resend the data packet after timing out on its wait for the acknowledgement from the receiver. This will continue for as long as there is a proper acknowledgement from the receiver and could take several seconds per packet.

Most end-users have devices such as cordless phones, microwave ovens and access points operating within WLAN environments, without having any idea about what threat they constitute to the sound functioning of the network. Some of these physical threats, which go unnoticed without the use of any tools based on frequency domain measurements, include:

Frequency-hopping devices: Frequency-hopping spread-spectrum (FHSS) devices generally spread their transmission spectrum across a wide range of frequencies and transmit by hopping over different frequency bands within the wide range many times a second. The earlier 802.11 standard was based on the FHSS, with transmission hopping across the entire 2.4-GHz frequency range several times each second . The entire 2.4 GHz was divided into 79 channels, each 1-MHz wide. This made jamming or interrupting FHSS transmissions difficult, while, at the same time, designing FHSS transmitters was relatively inexpensive.

Similarly, the Bluetooth protocol uses frequency hopping for data transmission, in the 2.4-GHz range, with 1,600 frequency hops per second over a bandwidth of 1 MHz. Cordless phones operating in the 2.4-GHz and 5-GHz frequency ranges also use frequency hopping.

Microwave ovens: Conventional microwave ovens operate within the 2.4-GHz ISM band. Microwave ovens are known to emit 1,000-1,200 watts of power within the heating chamber. Due to poor fabrication and RF shielding, however, RF energy can leak out of a microwave oven. As much as 10-dBm to 20-dBm power can be detected a couple of feet outside a microwave oven, sufficient enough to completely shut down the operation of a WLAN. In most cases, the operation of a microwave oven can bring down performance by 75%-90% if the network operates in the same channel on which the microwave ovens works.

A spectrum analyzer shows the relationship between the frequency and the power. It looks similar to an oscilloscope and has controls for varying the parameters of the spectrum of interest, such as the spectral span, center frequency, resolution bandwidth and reference level. Modern spectrum analyzers based on fast file transfer technology are lightweight, albeit specific in their application.

The presence of such interfering devices as cordless telephones, microwave ovens, Bluetooth devices and baby monitors can only be identified using frequency domain measurements. The presence of such devices can be attributed to an increase in the packet error rate obtained in the time domain measurements, but this is not always the case. By observing the spectrum, the exact nature, and subsequently, the type of interference device can be identified.

The accompanying illustration shows the current sweep waveform in green. The part of the current waveform marked as (1) represents a direct sequence spread-spectrum waveform and the transmitting device is, in this case, an 802.11b access point. A narrow-band transmission is represented in (3). To the inexperienced spectrum analyst, a narrow-band transmitter might seem to be operating close to an 802.11b access point. This is where the waveform peak hold feature comes in handy, making the discerning of FHSS waveform from narrow-band transmission easier.

The yellow trace in the illustration shows the peaks of the swept waveform held over a period of a number of sweeps. The waveform it represents is a FHSS transmitter–a Bluetooth device in this case. As can be seen, the transmitter transmits across the entire 50-MHz span of the current spectrum. The overlapping of the FHSS waveform and the DSSS waveform indicates severe interference of the Bluetooth device with the access point.

Using a continuous spectrum sweep serves as a good technique to locate wireless interference. Some RF transmission might be missed by the spectrum sweep, particularly if that span is large and the frequency resolution is poor. A good method would be to trigger the spectrum measurements only when the input power exceeds a certain threshold, preferably user input. This way, important wireless activity would be captured and displayed instead of the inconsequential noise floor.

Mapping of interference coverage can be an important tool in alleviating WLAN coverage problems. Large office spaces might have numerous non-802.11 sources of interference. With the help of an RF interference-mapping tool, an estimate can be made as to where the interference might cause a significant reduction in throughput. Plotting regions of interference can help in planning the deployment of a WLAN in order to optimize coverage and throughput. These analysis tools are helpful in understanding the interference contour of a facility where a WLAN has to be deployed.

Sandeep Natekar is a software engineer for Berkeley Varitronics Systems, Metuchen, N.J.

For more information:
www.rsleads.com/610cn-251