The project is concerned on research on a new radio receiver architecture with reduced analog receiver front-end. The main idea of this project is to remove interference from radio signals using compressed sensing so that extensive signal processing in the analog domain and high signal sampling rate can be avoided. Hence the direct conversion receiver complexity is pushed from the analog to the digital domain. This project may potentially lead to better performing and more power efficient receivers, which are more feasible to be implemented in integrated circuits. This 3-year project consists of both a theoretical and a practical part. The project is performed at Aalborg University, Denmark in cooperation with international academic institutions.
Problem and proposed solution
During the last twenty years superheterodyne receiver architecture has lost its position of the most popular architecture in favor of a direct conversion architecture. It is mostly due to growing popularity of mobile communication devices – direct conversion receivers are far more adequate in mobile applications due to flexibility and implementation costs.
If compared to the heterodyne receivers, direct conversion receivers are sensitive to interference. Due to still rising wireless traffic and increasing congestion, there is an interference problem faced by wireless communication engineers. Currenlty, the problem is addressed by adding a massive analog filtering in a receiver’s front-end to suppress the interference. Figure on the left shows an example of analog front-end of a direct conversion receiver. Integrated circuits implementation of the needed analog elements (mostly filters) is problematic due to occupied chip area and noise-coupling problem in analog integrated circuits.
Without the massive radio filtering, the radio signal would have to be sensed with an extremely high sampling frequency, which is not practically feasible. Compressed sensing is a relatively new signal processing paradigm, which may be used to address the above problem. Compressed sensing is still a relatively new discovery – so far it has mainly been used in image processing. Very limited research activities were undertaken to use compressed sensing in practical applications in wireless technology.
Problem: To reduce interference, wireless receivers require high quality analog filters, which are expensive and difficult in an integrated circuit implementation.
Proposed solution: Using computational power available nowadays in mobile devices to reduce interferences by digital signal processing.
Phases of the project
The project is planned for 3 years, in a time period 15th May 2013 – 14th May 2016. It consists of three phases which span unevenly over time.
- Scientific simulation [15 months] A system, which simulates the receiver and the interfered environment conditions will be developed. In this phase, the method of interference reduction will be developed on simulation level. This is a purely theoretical/simulation phase.
- Circuit-level simulation. [9 months] The developed interference mitigation method will be implemented in a simulated electronic circuit. Both integrated circuit and discrete circuit implementations will be simulated. This is a mixed practical/theoretical phase.
- Tests on the real circuits using discrete elements [12 months] The developed idea will be tested on a circuit that will be built with discrete elements. The circuit will be tested in both ideal and real environments.
The more detailed plan for the phases will be published shortly. In the foreseen outcome of the project there are three main contributions. The foreseen outcome of the project, in expected chronological order of completion, is as follows:
- A digital-processing-based method of interference reduction with ability of real-time exection.
- A simulated model of a circuit, which realizes the developed method
- A prototype circuit based on discrete elements, which practically demonstrates the developed concept.
The above steps of the outcome are strongly connected to the phases of the project.
The project comprises 3 hypotheses. The hypotheses are as follows:
- With compressed sensing it is viable to reduce interference in mobile communication systems without using the massive analog filtering of radio signals.
- It is possible to design a hybrid reconstruction algorithm, dedicated solely to a concerned problem, which ensures very good signal reconstruction quality and may be implemented in a real-time execution regime.
- The proposed solution will not increase the power dissipated by the chip and significantly decrease usage of the chip area.
First two of them refer to the theoretical/simulation research phases of the project. The last refers to the practical part of the project and, to a certain extend, depends on technological constraints.