Method and system for compressed sensing joint channel estimation in a cellular communications network
Methods and systems for performing compressed time domain joint channel estimation in a multi-user MIMO wireless network include receiving data corresponding to transmission of training signals from a plurality of users to a base station over a MIMO wireless network, determining a limited data set by limiting the received data in a time domain according to an estimated maximum delay spread, forming a well-conditioned low rank training matrix by identifying a channel model, estimating an active tap from the formed well-conditioned low rank training matrix, and subtracting a contribution of the selected active tap from the limited data set.
1. A method of performing time domain channel estimation in a multi-user multiple input multiple output (“MIMO”) wireless network, the method comprising:
receiving data corresponding to transmission of training signals from a plurality of users to a base station over a MIMO wireless network;
determining a limited data set by limiting the received data in a time domain according to an estimated maximum delay spread;
forming a well-conditioned low rank training matrix by identifying a channel model;
estimating an active tap from the formed well-conditioned low rank training matrix; and
subtracting a contribution of the selected active tap from the limited data set.
2. The method of claim 1 , wherein estimating an active tap from the formed well-conditioned low rank training matrix comprises:
selecting an active tap from the limited data set; and
determining a number of active taps in the MIMO multi-user wireless network.
3. The method of claim 2 , wherein the active tap is selected by at least one of orthogonal matching pursuit and basis pursuit.
4. The method of claim 2 , wherein selecting an active tap includes selecting a plurality of active taps.
5. The method of claim 4 , wherein the plurality of active taps is selected according to which active taps have the strongest correlation to the received training signal.
6. The method of claim 4 , wherein for each user the plurality of active taps includes at least one active tap corresponding to that user.
7. The method of claim 2 , wherein selecting an active tap from the limited data set comprises:
detecting active tap locations of the reduced data set; and
selecting the active tap according to the estimated active tap locations.
8. The method of claim 7 , wherein the active tap locations are detected using a matching pursuit algorithm.
9. The method of claim 7 , wherein the active tap is selected by solving a l1/l2 norm minimization problem.
10. The method of claim 2 , wherein selecting an active tap from the limited data set comprises applying a basis pursuit de-noising function associated with an l1-norm objective function to the reduced data set.
11. The method of claim 2 , comprising repeating the steps of selecting and subtracting until a residual signal norm falls below a specified minimum.
12. The method of claim 1 , wherein at least one of the training signals is a sounding reference signal (“SRS”) signal.
13. The method of claim 1 , wherein at least one of the training signals is a demodulation reference signal (“DMRS”) signal.
14. The method of claim 1 , wherein the MIMO wireless network is a cloud radio access network (“C-RAN”) network.
15. The method of claim 1 , wherein the network operates using at least one of time division duplexed (“TDD”) and frequency division duplexed (“FDD”) communications.
16. The method of claim 1 , wherein the active tap is estimated using at least one of:
l1-norm minimization;
l2-norm minimization;
regulated L2-norm immunization;
OMP greedy matching pursuit; and
stomp greedy matching pursuit.
17. A system for performing time domain channel estimation in a multi-user multiple input multiple output (“MIMO”) wireless network, the system comprising:
a signal receiving unit configured to receive data corresponding to transmission of training signals from a plurality of users to a base station over a MIMO wireless network; and
a signal processing unit configured to
receive data corresponding to transmission of training signals from a plurality of users to a base station over a MIMO wireless network;
determine a limited data set by limiting the received data in a time domain according to an estimated maximum delay spread;
form a well-conditioned low rank training matrix by identifying a channel model;
estimate an active tap from the formed well-conditioned low rank training matrix; and
subtract a contribution of the selected active tap from the limited data set.
18. The system of claim 17 , wherein the signal processing unit is further configured to estimate an active tap from the formed well-conditioned low rank training matrix by:
selecting an active tap from the limited data set; and
determining a number of active taps in the MIMO multi-user wireless network.
19. The system of claim 18 , wherein the signal processing unit is further configured to repeatedly select the active tap and subtract the contribution until a residual signal norm falls below a specified minimum.
20. The system of claim 18 , wherein the signal processing unit is configured to select the active tap by at least one of orthogonal matching pursuit and basis pursuit.
21. The system of claim 18 , wherein the signal processing unit is configured to select a plurality of active taps.
22. The system of claim 21 , wherein the signal processing unit is configured to select the plurality of active taps according to which active taps have the strongest correlation to the received training signal.
23. The system of claim 21 , wherein the signal processing unit is configured to select the plurality of active taps such that, for each user, the plurality of active taps includes at least one active tap corresponding to that user.
24. The system of claim 18 , wherein the signal processing unit is configured to select and estimate the active tap by:
determining a number of active taps in the wireless network;
identifying a channel model;
forming a well-conditioned low rank training matrix; and
estimating the active tap from the low rank training matrix.
25. The system of claim 18 , wherein the signal processing unit is configured to select the active tap by:
detecting active tap locations of the reduced data set; and
selecting the active tap according to the estimated active tap locations.
26. The system of claim 25 , wherein the signal processing unit is configured to detect the active tap locations using a matching pursuit algorithm.
27. The system of claim 25 , wherein the signal processing unit is configured to select the active tap by solving an l1/l2 norm minimization problem.
28. The system of claim 18 , wherein the signal processing unit is configured to select the active tap by applying a basis pursuit de-noising function associated with an l1-norm objective function to the reduced data set.
29. The system of claim 17 , wherein at least one of the training signals is a sounding reference signal (“SRS”) signal.
30. The system of claim 17 , wherein at least one of the training signals is a demodulation reference signal (“DMRS”) signal.
31. The system of claim 17 , wherein the MIMO wireless network is a cloud radio access network (“C-RAN”) network.
32. The system of claim 17 , wherein the network operates using time division duplexed (“TDD”) and frequency division duplexed (“FDD”) communications.
33. The system of claim 17 , wherein the signal processing unit is further configured to estimate the active tap using at least one of:
l1-norm minimization;
l2-norm minimization;
regulated L2-norm immunization;
OMP greedy matching pursuit; and
stomp greedy matching pursuit.
34. A non-transitory computer readable medium storing a computer program, executable by a machine, for performing time domain channel estimation in a multi-user multiple input multiple output (“MIMO”) wireless network, the computer program comprising executable instructions for:
receiving data corresponding to transmission of training signals from a plurality of users to a base station over a MIMO wireless network;
determining a limited data set by limiting the received data in a time domain according to an estimated maximum delay spread;
forming a well-conditioned low rank training matrix by identifying a channel model;
estimating an active tap from the formed well-conditioned low rank training matrix; and
subtracting a contribution of the selected active tap from the limited data set.