Design of a Filter

Hi,

I am modelling the physical layer of an 802.11 transmitter (to
implement it into an FPGA). At this moment I want to design the digital
filter responsible of pulse shaping (before the DAC). My doubts are:

1. FIR or IIR filter?, why?.
2. For each option (FIR or IIR) there are different types of filters
(FIR: Raised_cos_Barlett, Raised_cos_Blackmann, etc, ... or IIR:
Butterworth, Chebyshev, Bessel, ...) so, which should I choose?, why?.

I enclose the requisites of the standard:

The  "transmit spectrum mask" , is defined as:
----------------------------------------------------------------------------------------


The transmitted spectral products shall be less than -30 dBr (dB
relative to the SINx/x peak) for

        fc-22Mhz <f < fc-11Mhz; and
        fc+11Mhz <f< fc+22Mhz;

and shall be less than -50 dBr for

        f<fc-22Mhz; and
        f>fc+22Mhz

where fc is the channel center frequency.

--------------------------------------------------------------------------------------------

Regards.

Jajo