INTRODUCTION

1.1   Receiver Architectures

         The traditionally used architecture, the  superheterodyne or superhet in short               

      (Fig.   1.1),   down-converts   a   RF  signal  to  baseband  in  one  or  sometimes  two  stages

  depending  on  the  frequency  plan  for the  chosen  architecture.  Mixers, used for this down

  conversion,    convert   the  RF  frequency  to   a  lower  intermediate  frequency   (IF)  for

filtering  and  amplification  before  finally  converting  it  to  baseband  .  The  IF is defined as:

(1.1)                                                                                                        

Where    and    are the frequency  of  the  RF  and local  oscillator  (LO)  signals 

respectively.  These signals  form  the  input  to the  mixer, which  performs  analog

multiplication   on  them  to  give  rise  to  the  signal  at IF.  The major   drawback   of   this

 architecture  which  also  limits  its  integrability  onto  a single  chip  is  the  presence  of  image 

frequencies.  An  image  is  defined  as  a  signal  other  than  the  signal of  interest  that  mixes 

to   the  same  IF  as  the  desired  signal.  Down-conversion  of  two  different  frequencies  to

the  same  IF  occurs because  of  the  inability  of the  mixer  to  recognize  the  polarity  of  the

frequency  difference  between  the  RF  and  LO.  For  example, if  the  RF  signal  is  located

one  IF  higher  than  the  LO  (low-side  injection),  the  image  frequency  is  located  at,

                                                                                  (1.2)             

And  if  the  RF  signal  is  located  one  IF  lower  than  the  LO  (high-side  injection),  image

Frequency  is  located  at,

                                               (1.3)

Therefore in the superhet architecture the mixer has to be preceded by filters for the

Purpose of image rejection and channel selection and these filters have to have steep

roll-off and very high out-of-band rejection to attenuate the unwanted signals.  To

acoommodate  these stringent requirements, the filters need to be designed with high

Quality (Q) factor resonators and multiple poles. The low Q factor of on-chip inductors

results in prohibitively high passband insertion loss for multiple-poled integrated

inductor and capacitor (LC) filters. Furthermore, since monolithic  inductor and

capacitors require a great deal of die area, multiple pole LC filters quickly become

excessively large for on-chip integration.

                                       Figure 1.1 Superheterodyne receiver architecture[2]

1.1.1  Direct Conversion Receivers (DCRs)

            An alternative approach to heterodyning described above is to convert the RF

Frequency directly to baseband frequency thereby getting rid of multiple stages of