Spectrally efficient emission mask shaping for OFDM cognitive radios

Orthogonal Frequency Division Multiplexing has been widely adopted in recent years due to its inherent spectral efficiency and robustness to impulsive noise and fading. For cognitive radio applications in particular, it can enable flexible and agile spectrum allocation, yet suffers from spectral leakage in the form of large side lobes, leading to inter-channel interference, unless mitigated carefully. Hence, recent OFDM-based standards such as 802.11p for vehicular communication and 802.11af for TV whitespace impose strict spectrum emission mask limits to combat adjacent channel interference. Stricter masks allow channels to operate closer together, improving spectral efficiency at the cost of implementation difficulty. Meeting the strict limits is a significant challenge for implementing both 802.11p and 802.11af, yet remains an important requirement for enabling cost-effective systems. This paper proposes a novel method that embeds baseband filtering within a cognitive radio architecture to meet the specification for the most stringent 802.11p and 802.11af masks, while allowing ten 802.11af sub-carriers to occupy a single basic channel without violating SEM specifications. The proposed method, performed at baseband, relaxes otherwise strict RF filter requirements, allowing the RF subsystem to be implemented using much less stringent 802.11a designs, allowing cost reductions.