Crest factor reduction in an OFDM/ WiMAX network

Abstract

Broadband services require multicarrier modulation schemes such as Orthogonal Frequency Division Multiplexing (OFDM) because they are robust to channel hostilities and support high data rates. These schemes present a problem of a high crest factor which puts heavy demand on the transmitter High Power Amplifier to have a large linear dynamic range thus increasing the cost of the system. A crest factor reduction technique was developed that added an additive signal to the OFDM signal that canceled out the signal peaks. It only added the additive signal when the crest factor of the OFDM signal was above a threshold crest factor. Two forms of the developed crest factor reduction technique were presented depending on whether peak re-growth was controlled or not. The developed crest factor reduction technique was simulated for a Worldwide Interoperability for Microwave Access (WiMAX) network. The technique with peak re-growth gave an average Peak to Average Power Ratio reduction of 0.95 decibels (dB) with standard deviation of ±0.5 dB while the technique that controlled peak re-growth gave 1.7 dB with standard deviation of ±0.3dB. The technique which controlled peak re-growth gave better crest factor reduction. Three crest factor reduction techniques were studied and simulated. These were: selected mapping, selected tone reservation and tone reservation. They gave an average Peak to Average Power Ratio reduction of one decibel which was less than that for the developed technique that controlled peak re-growth. It was established that the developed technique without peak re-growth gave optimum performance when 16 mappings were used; the threshold was set at six decibels and with unity gain applied to the additive signal. An optimum performance referred to the point at which the gap between the Peak to Average Power Ratio reduction and the increase in transmit power was at its largest.