TV White Space Network Planning and Co-Channel Interference Estimation
Permanent address of the item is
The frequencies that are suitable for radio communication are a scarce resource and some of the licensed frequency bands are considered to be underexploited. The future visions comprehend using the spectrum more efficiently, but avoiding the interferences by the means of cognitive radios. A cognitive radio is a device that can observe the radio conditions of its surrounding environment, learn from this and use this information for adjusting its operation to the current state of the environment. One of the licensed frequency bands that have been opened for cognitive radio testing is the terrestrial TV band. The regionally or temporally free channels in the TV band are called the white space. This thesis discusses the white space network planning and the co-channel interference estimation. The co-channel interference is due to the Digital Terrestrial Television (DTT) network. Especially this thesis has an emphasis on the rural wireless broadband application. The main focus is on a pilot case providing a fixed wireless broadband to a few locations in Kirkkonummi using the Espoo TV and radio mast. The starting point for the network planning is determining a free channel so that no interference is caused to the primary user of the frequency band, that is, the DTT network. After selecting the free channel, the downlink coverage and capacity is estimated with the network planning tool Atoll. Some of the white space studies made so far have indicated that the uplink connection limits the coverage of the network due to the co-channel interference in the base station antenna from the DTT transmitters. In order to investigate this issue, a Matlab estimation function to estimate the co-channel interference power will be introduced in this thesis. The estimation function is based on the ITU-R P.1546 propagation model. The estimation function will be evaluated with the interference power measurements from three different locations, from Jokela, Pasila and Espoo. The co-channel interference will be used to estimate the uplink coverage in the Espoo pilot and also to calculate the amount of usable channels based on the interference level. The network planning, added with the uplink coverage estimation, will be evaluated with field tests. The most important conclusion in this thesis is that estimating the uplink limitation is an essential part of the white space network planning even with point-to-point connections. The estimation has its own challenges, but rather good results can be achieved with the existing propagation models with small changes. Furthermore it was found that using such devices in the white space band that are not designed for such purposes is challenging. In addition to the required cognitive capabilities, the white space devices demand a wide operation region to be able to use the best channels in the area. Also a good selectivity and strict emission mask are required to overcome the different interference scenarios in the white space band.