Despite their many advantages, OFDM systems have a few, but important, drawbacks. In particular, OFDM relies on the inverse FFT for modulation purposes, which leads to a very poor spectral containment and a high susceptibility to narrowband noise. To mitigate this problem, we propose in this thesis to perform multicarrier modulation using a perfect reconstruction (PR) DFT filter bank instead of employing the inverse FFT. The design of such filter banks is addressed using a novel method that guarantees the PR property to be satisfied while the spectral containment is being maximized. Equalization in the proposed DFT filter bank transceiver takes advantage of the fact that the filter banks do not contribute to any distortion due to its PR nature. Two equalization schemes are presented. The first one is based on a zero-padded block linear equalization approach, and the second one utilizes a one-tap per subcarrier configuration. The estimation of the channel coefficients in the proposed transceiver is also addressed. A blind estimation method that exploits the inherent cyclostationarity of the transmitted signal is derived. Computer experiments presented in this thesis indicate that the spectral containment of the proposed PR DFT filter bank transceiver is indeed superior to that of the OFDM system. Moreover, simulations conducted in a DSL-like environment contaminated by a narrowband noise show that the achievable bit rate of the proposed transceiver is much higher than that of a conventional OFDM system.