Wireless network, including wireless metropolitan area networks (WMAN) such as those compliant with the IEEE standard 802.16.x (WiMAX), may use multiple antennas on the transmitters and receivers, referred to as Multiple-Input Multiple-Output (MIMO), to communicate in order cancel interference from adjacent cells. Wireless networks may communicate using Orthogonal Frequency Division Multiplexing (OFDM) signaling. An OFDM signal is comprised of multiple sub-carriers each modulated at a symbol rate equal to the reciprocal of the frequency separation. MIMO schemes are often implemented with OFDM signaling as OFDM provides for easy characterizing of channel frequency response.
For wireless transmissions where high data rates and high signal to interference and noise ratio (SINR) are desired, a wireless MIMO receiver may operate in a spatial (de)multiplexing (SM) mode to estimate the transmitted signal. For wireless transmissions where increased coverage at low SINR is desired, the wireless MIMO receiver may operate in a space-time block (de)coding (STBC) mode. In order to match the transmission to the channel conditions, the MIMO receiver scheme may switch between SM and STBC modes (detection modes) and/or may adapt the number of received sub-streams (RF chains) depending on the operating power mode and channel conditions. The MIMO receiver may need to switch between the detection modes with minimum latency.
The MIMO receivers may include a plurality of different detectors (e.g., maximum ratio combining (MRC), minimum mean squared error (MMSE), maximum likelihood (ML)) to account for the different MIMO modes. The appropriate detector may be enabled based on the spatial operational MIMO mode. Having multiple MIMO detectors requires silicon area for each detector and may require complicated data interfaces.
The features and advantages of the various embodiments will become apparent from the following detailed description in which:
Maximum likelihood detectors (MLD) can be used in spatial (de)multiplexing (SM) mode to estimate the transmitted signal from the received signal. The MLD compares the received signal with all possible transmitted signals and estimates s according to its closest match. At the receiver, the most likely transmitted signal is identified as SmL=arg min(sjε{s1,s2, . . . , sK}){∥r−H sj∥2}, where the search for minima for a M-QAM constellation is conducted for all MNt possibilities in s, where Nt is the number of transmitter antennas in the MIMO system and M is the number of constellation points. The MLD may also deliver the reliability values associated the most likely transmitted signal, which are known as soft-decision or Log-likelihood ratio (LLR) outputs.
The MLD can be simplified by scanning the hypotheses for all transmitting antennas except one, and finding the remaining signal by applying maximum ratio combining (MRC) and slicing. Using the simplified soft output MLD (SMLD) reduces the search for possible constellations for s to MNt-1 (by a factor of 1/M). For a MIMO system having 2 transmitters (2×Nr MIMO) H=[h1 h2] where the column vectors h1 and h2 are the Nrx1 channel gain vectors corresponding to the 2 transmitted signal s1 and s2. The SMLD performs SM on a 2×Nr MIMO as follows.
All the possible constellation points for s1 are scanned. For each s1 hypothesis the ML solution of s2 is found by MRC and slicing, such that
s2s(r,s1)=slice{(∥h2∥2)−1h2Hr−(∥h2∥2)−1h2Hh1s1} (Equation 1A)
The Euclidean distances d1 for each s1 can be calculated as
Note that the terms t1a-t1e are used to name the product terms they fall below.
Note that the energy of the received signal ∥r1∥2 is crossed out because the value is same for each s1 and need not be considered. For each bit bi of stream s1, the d1 is partitioned as
d1+={d1}|b
The LLR of each bi is calculated as
LLR(bi)|s1=min{d1+}−min{d1−} (Equation 4A)
The same process is followed for all the possible combinations of s2
s1s(r,s1)=slice{(∥h1∥2)−1h1Hr−(∥h1∥2)−1h1Hh2s2} (Equation 1B)
d
2+
={d
2}|b
LLR(bi)|s2=min{d2+}−min{d2−} (Equation 4B)
The SMLD provides improved packet error rates (PER) for frequency selective channels, especially in presence of mutual interference. The SMLD scheme described above for a 2×NrMIMO SM mode may be utilized for other detection modes.
In a 2×Nr MIMO STBC mode having an orthogonal H, ∥h1∥2=∥h2∥2 and h1Hh2=0. Therefore the maximum likelihood of for s1 is independent of and s2 and vice versa, such that
s2s(r,s1)=slice{(∥h2∥2)−1h2Hr}; s1s(r,s2)=slice{(∥h1∥2)−1h1Hr} (Equations 1C, 1D)
Since the MRC contribution is independent of search variable, the MRC calculations of the SMLD are not required. Accordingly the Euclidean distance calculations can exclude common and uncorrelated values, such that
d1(s,s2s)=∥r∥2+∥h1∥2|s|2+∥h2∥2|s2s|2−2 Re{(rHh1)s}−2 Re{(rHh2)s2s}+2 Re{s1H(h1Hh2)s2s} (Equation 2C)
d2(s1s,s)=∥r∥2+∥h1∥2|s1s|2+∥h2∥2|s|2−2 Re{(rHh1)s1s}−2 Re{(rHh2)s}+2 Re{s1sH(h1Hh2)s2} (Equation 2D)
While the SMLD framework may be the same for SM and STBC detection modes, the amount of computation that is actually utilized for the STBC MIMO is much less. Dividing the SMLD computations into discrete operations and enabling the appropriate operations based on mode enables a single SMLD to be utilized for both SM and STBC detection modes. Using the SMLD for STBC where computations not required can be skipped makes using SMLD for STBC an efficient option since excess operations will not be performed. Using SMLD for STBC results in improved packet error rates compared to other detectors typically used for STBC (e.g., MRC).
The number of operations that are activated depends on the MIMO configuration (Nt ×Nr) and the detection mode (e.g., SM, STBC). For example, WiMax OFDM access systems may have between 1-2 transmitters and 1-3 receivers. The MIMO systems (2×2, 2×3) may be operated in either SM or STBC detection mode. A controller within the MIMO receiver can set the SMLD for the appropriate configuration.
The unified SMLD eliminates the need for multiple detector engines in the receiver. This may reduce the die area used for detectors and may simplify the control structure. The unified SMLD would have high (e.g., 100%) utilization and would require only a single data interface. The unified SMLD may also lower power consumption.
The unified detector has been described with respect to an SMLD detector and SM and STBC detection modes but is not limited thereto. Rather, other type of detectors now known or later discovered may be utilized if the computational operations can form a common framework for various detection modes and/or MIMO configurations in which the computational operations performed are based on some combination of detection modes and MIMO configuration.
Although the disclosure has been illustrated by reference to specific embodiments, it will be apparent that the disclosure is not limited thereto as various changes and modifications may be made thereto without departing from the scope. Reference to “one embodiment” or “an embodiment” means that a particular feature, structure or characteristic described therein is included in at least one embodiment. Thus, the appearances of the phrase “in one embodiment” or “in an embodiment” appearing in various places throughout the specification are not necessarily all referring to the same embodiment.
The various embodiments are intended to be protected broadly within the spirit and scope of the appended claims.
Number | Name | Date | Kind |
---|---|---|---|
6470043 | Lo et al. | Oct 2002 | B1 |
7146134 | Moon et al. | Dec 2006 | B2 |
20030235147 | Walton et al. | Dec 2003 | A1 |
20050075081 | Catreux-Erceg et al. | Apr 2005 | A1 |
20060067277 | Thomas et al. | Mar 2006 | A1 |
20080181094 | Lou et al. | Jul 2008 | A1 |
20090190682 | Ouyang et al. | Jul 2009 | A1 |
Number | Date | Country | |
---|---|---|---|
20090034662 A1 | Feb 2009 | US |