Radio local area network (RLAN) and C-Band weather radar interference studies

There is a growing concern about the effect that wireless internet devices operating at 5 GHz will have on the operation of weather radars. At the World Radio-communication Conference (2003) (WRC-03), the decision was made to provide a primary allocation to the mobile service for the implementation of wireless access systems, including radio local area networks, in the band 5470-5725 MHz. The operation of RLANs in this band is permitted as long as they do not cause interference to licensed services such as Radiolocation, which operates in the 5600-5650 MHz sub-band (C-Band).

RLAN's transmit packets at various modulations, pulse rates and carrier frequencies, which extend beyond the weather radar C Band frequencies (5470-5725 MHz). RLANs would be able to use any channalization scheme within the new bands of 5470MHz to 5725MHz, adding to the existing bands of 5250MHz to 5350MHz and 5725MHz to 5825MHz already in operation in Canada RLAN interfernece appear as additive noise to the weather radar. To mitigate potential interference, new regulations stipulate the use of Dynamic Frequency Selection (DFS). Before using a channel, the RLAN must check for the presence of radars for a 60-second period. Once the RLAN is using a channel, it must continue to monitor for the presence of radar signals. If these signals are detected, RLANs must vacate the channel for a 30-minute period. In addition, before re-using the channel, the RLAN must continuously monitor the channel for a 10-minute period. Or, the RLAN can notch out the sub-band 5600-5650 (channel exclusion). These latter requirements were put forward by Canada to protect its weather radars and were adopted by the WRC-03.

Field experiments using real RLAN sources and a C Band weather radar were conducted where (i) a RLAN signal was directly injected into the receive port of the weather radar to verify the additive noise nature of the interference on the weather radar signal, and (ii) a RLAN signal illuminated an operational weather radar as it was pointed in various directions. The first tests confirmed the additive noise behaviour of the RLAN interference. The latter tests indicated various modes of interference are possible - for example (when the RLAN was located in close proximity to the radar) the interfering signal could not only be seen directly along the boresight of the antenna but also off-axis through backscatter and reflection from the vegetation. These initial tests allowed for the simulation of RLANs in a rural-urban environment where various distributions of the RLANS are located in various propagations environments. The conclusion is that the DFS and a 10 minute minute monitoring requirement are needed to mitigate the interference.