MIMO Communications with Limited Feedback
Why Feedback and Limited Feedback?
In a MIMO wireless communication system, channel sate information (CSI) at transmitters allows adaptive transmission, reduces receiver complexity, facilitates scheduling and enables multiple access, which leads to higher system throughput, longer battery lives, and fairer sharing of communication resources between users. Some examples are given as follows. In a precoding system, CSI allows a transmitter to send data along the strongest eigenmodes of a channel and effectively cope with fading. Transmit CSI also enables spatial division multiple access (SDMA), which supports multiple users without requiring additional bandwidth. Furthermore, transmit CSI allows opportunistic scheduling of users with good channels and thus increases system throughput.
For MIMO channels, efficient transport of CSI from receivers to transmitter is very challenging given the multiplicity of channel coefficients, the small bandwidth of a feedback control channel, and the fact that this channel is shared by users in many communication systems. Therefore, efficient CSI feedback demands high-ratio compression of feedback CSI as well as its optimal employment in transmission and scheduling. The related research area is known as limited feedback, which has been very active recently due to its importance for implementing modern communication systems including WiMax and 3GPP-LTE.
What
do we do?
At WSIL, we have been carrying out extensive research on limited feedback mainly following three approaches. First, various limited feedback methods are being designed using subspace quantization techniques and other tools from communication theory, signal processing, and applied mathematics. These limited feedback methods are targeted at various MIMO systems including space-time block codes, beamforming, and spatial multiplexing systems. These methods are also being extended to broad band and non-stationary MIMO channels. As the result, algorithms for MIMO-OFDM and multi-mode limited feedback have been designed. Second, addressing the multi-user scenario, the methods for scheduling based on limited feedback and distributed feedback control are also being developed. Third, we investigate the impact of various practical factors such as channel temporal correlation and feedback delay on limited feedback delay and performance.
What
have we done?
Our group has demonstrated close-to-optimal performance for limited feedback precoding as well as multi-mode antenna selection techniques. We have introduced Grassmannian Subspace Packing based limited feedback methods and have shown their effectiveness for both narrow and broad band MIMO. We have developed efficient algorithms for feedback compression in the time, frequency and spatial dimensions. Moreover, we have developed limited feedback SDMA algorithms and threshold based feedback methods for constraining the multiuser feedback rate. Follow the links to the left for more details in our research and related publications.
This material is based upon work supported by the National Science Foundation under Grant No. 0514194, the Office of Naval Research under grant number N00014-05-1-0169, Freescale and Motorola. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation, the Office of Naval Research, Freescale or Motorola.
