Compressed Sensing Methods in Radio Receivers Exposed to Noise and Interference

Currently we observe a dramatical increase of wireless data trac and a rising number of mobile radio transceivers like Bluetooth, WiFi or LTE. Wireless
networks become more robust and deliver data rates many times higher than
rates available a couple of years ago. LTE, which is the currently implemented standard for mobile communication, implements performance parameters which not so long ago were available only for cable networks.
There is also observed a trend to implement many radio receivers in a
single integrated circuit. Radio receivers become more software-based, due
to signicant increase of computational power avaialible in integrated circuits.
Nevertheless, some necessary analog-domain processing is still a must. Analog
elements are an obstacle in integrated implementation of radio receivers, digitization of radio receivers is desirable. Unfortunatelly, due to the constantly
increasing trac, there is a problem of interference, which makes digitization
of radio receivers even more dicult. High-order low-pass lters are needed
to remove interfering signals and secure a high-quality reception.
In the mid-2000s a new method of signal acquisition, called compressed
sensing, emerged. Compressed sensing is a mathematical tool which allows
for sub-Nyquist signal sampling. In this thesis the author opens a new possibility of relaxing requirements for analog signal ltering in a direct conversion
receiver by applying compressed sensing. In the proposed solution,high-order
low-pass lters which separate the downconverted baseband signal and interference, may be replaced by low-order lters. Additional digital signal processing is a price to pay for this feature. Hence, the signal processing is moved
from the analog to the digital domain.
Filtering compressed sensing, which is a new application of compressed
sensing, is proposed. The author shows that by applying appropriate signal
acquisition it is possible to lter out the wanted signal with a reduced signal
sampling frequency and reduced lter order. A compressed-sensing based
direct conversion receiver is proposed. An LTE signal generator, which is a
side accomplishment of the project is also presented.