Search results

The rapidly increasing popularity of WiFi has created unprecedent levels of congestion in the unlicensed frequency bands, especially in densely populated urban areas. This results mainly because of the uncoordinated operation and the unmanaged interference between WiFi access points. Recently, Radio Environment Maps (REM) have been suggested as a support for coordination strategies that optimize the overall WiFi network performance. Despite some theoretical work done in this area, there are no clear experimental evidences of the benefit brought by WiFi coordination. In this context, the main objective of this experiment is to assess the benefit of a coordinated management of radio resources in dense WiFi networks using REMs for indoor scenarios. This experiment has used the w-iLab.t test environment provided by iMINDS, a cognitive-radio testbed for remote experimentation. It was shown that REMs are capable of detecting the presence of interfering links on the network (co-channel or adjacent channel interference), and a suitable coordination strategy can use this information to reconfigure Access Points (AP) channel assignment and reestablish the client connection. The coordination strategy almost double the capacity of a WiFi link under strong co–channel interference, from 6.8 Mbps to 11.8 Mbps, increasing the aggregate throughput of the network from 58.7 Mbps to 71.5 Mbps. However, this gain comes with the cost of a relatively high density network of spectrum sensors (12 sensors for an area of 60 × 20 m), increasing the cost of deployment.

The rapidly increasing popularity of WiFi has created unprecedented levels of congestion in the unlicensed frequency bands, especially in densely populated urban areas. This results mainly because of the uncoordinated operation and the unmanaged interference between WiFi access points. In this context, the main objective of this experiment is to assess the benefit of a coordinated management of radio resources in dense WiFi networks for both 2.4 GHz and 5 GHz bands, using Radio Environment Maps (REM). This experiment has used the w-iLab.t test environment and the portable test-bed provided by iMINDS for indoor scenarios. It was shown that REMs can detect the presence of interfering links on the network (co-channel or adjacent channel interference), and a suitable coordination strategy can use this information to reconfigure Access Points (AP) channel assignment and re-establish the client connection. The coordination strategy almost double the capacity of a WiFi link under strong co–channel interference, from 6.8 Mbps to 11.8 Mbps, increasing the aggregate throughput of the network from 58.7 Mbps to 71.5 Mbps. However, this gain comes with the cost of a relatively high-density network of spectrum sensors, increasing the cost of deployment. The technique of AP handoff was tested to balance the load form one AP to another, although the aggregate throughput is lower after load balancing. REMs are also capable of detecting coverage holes on the network, and a suitable Radio Resource Management strategy use this information to reconfigure the APs transmit power to reestablish the client connection and increase the throughput of the overloaded AP, at a cost of diminishing the aggregate throughput of the network. The insights coming out from this experiment helped to understand the opportunities and limitations of WiFi coordination strategies in realistic scenarios.