The present invention pertains to the field of cellular communication. More particularly, the present invention pertains to MIMO processing in DS-CDMA.
It is known that under suitable channel fading conditions, having both multiple transmit and multiple receive antennas—i.e., a MIMO (multiple input multiple output) channel—provides additional spatial dimensions for communication.
In a DS-CDMA (Direct Sequence—Code Division Multiple Access) communication system, orthogonal codes are used in order to arrange that a single carrier provides multiple access. In a MIMO 3GPP (Third Generation Partnership Project) transmission, a single (high-rate) stream is transmitted as a plurality of low-rate streams each using codes from a same set of codes in the HS-DSCH (High Speed-Downlink Shared Channel) and some of the codes are possibly also used—i.e. “reused”—by more than one antenna (but not others of the codes). In some such transmissions, the SINR (signal to interference plus noise ratio) for each code (and so the corresponding so-called code channel) of the HS-DSCH channel may differ.
What is needed is a way of taking this difference into account, and preferably in a way that is of use not only in a telecommunication per 3GPPP, but in any cellular telecommunication system having MIMO DS-CDMA functionality.
Accordingly, in a first aspect of the invention, a method is provided comprising: a step in which channel coding is performed for communication of systematic bits via a communication channel of a cellular communication system, the channel coding providing a stream of coded bits including both the systematic bits and also redundancy bits; and a step of allocating the systematic and redundancy bits to a plurality of code channels including both reused and non-reused code channels, each code channel corresponding to a different physical channel; wherein the bit allocation takes into account differences in the quality of the different physical channels by using a bit allocation rule that allocates more systematic bits to non-reused code channels than to reused code channels.
In accord with the first aspect of the invention, the stream for which the channel coding is performed may be a stream in a plurality of streams that in combination convey a single higher-rate data stream, and the bit allocation may be performed for each of the streams in the plurality of streams. Further, the bit allocation may provide bits for code channels in the plurality of code channels not reused by any of the streams, and for code channels in the plurality of code channels reused by two of the streams, and for code channels reused by three of the streams, and so on.
Also in accord with the first aspect of the invention, the method may further comprise a physical layer HARQ processing step subsequent to the step of channel coding, and the differences may be taken into account in the physical layer HARQ processing step.
Also in accord with the first aspect of the invention, the method may further comprise an interleaving step subsequent to the step of channel coding, and the differences may be taken into account in the interleaving step.
Also in accord with the first aspect of the invention, the method may further comprise a physical channel segmentation step subsequent to the step of channel coding, and the differences may be taken into account in the physical channel segmentation step.
Also in accord with the first aspect of the invention, if there are more systematic bits than can be allocated to non-reused code channels so that non-allocated systematic bits remain after allocating as many of the systematic bits as possible to non-reused code channels, the rule may allocate as many of the non-allocated systematic bits to the reused code channel expected to have the highest channel quality of the reused code channels.
In a second aspect of the invention, a computer program product is provided comprising a computer readable storage structure embodying computer program code thereon for execution by a computer processor, wherein the computer program code comprises instructions for performing a method according to the first aspect of the invention.
In a third aspect of the invention, an apparatus is provided comprising means for performing the steps of a method according to the first aspect of the invention.
In a fourth aspect of the invention, a user equipment terminal is provided having a transceiver for coupling the user equipment terminal to a radio access network, and including an apparatus comprising means for performing the steps of a method according to the first aspect of the invention.
In a fifth aspect of the invention, a network element is provided serving as at least part of a service access point of a radio access network, and including an apparatus comprising means for performing the steps of a method according to the first aspect of the invention.
In a sixth aspect of the invention, a system is provided, comprising: a radio access network including: network element serving as at least a part of a service access point of the radio access network, for providing a cellular communication signal; and a plurality of user equipment terminals each responsive to at least a portion of the cellular communication signal, wherein at least one of either the network element or one or more of the user equipment terminals includes an apparatus according to the third aspect of the invention.
The above and other objects, features and advantages of the invention will become apparent from a consideration of the subsequent detailed description presented in connection with accompanying drawings, in which:
The invention is here described in case of communication of a MIMO transmission via HS-DSCH of a 3GPP cellular telecommunication system. It should be understood however that the invention is of use in case of a multi-streaming MIMO transmission communication via a communication channel of any cellular communication system having MIMO DS-CDMA functionality.
As already specified (in 3GPP TS25.212 v.5.4.0 2003-03) for single input single output (SISO) transmission in HS-DSCH, the systematic bits in a turbo code may be assigned/allocated to more reliable locations in a 16-QAM constellation. In addition, the bit allocation may be rotated during a HARQ (hybrid automatic repeat request) process in case of failed packets. The invention uses these ideas in a MIMO transmission in an unbalanced code re-use scenario.
In coding for HS-DSCH for SISO, data arrives at the coding unit in the form of a maximum of one transport block once every transmission time interval (TTI). The transmission time interval is 2 ms in duration, and is mapped to a radio sub-frame of 3 slots. As shown in
In the channel coding step, which for the HS-DSCH transport channel usually uses a rate ⅓ turbo coder, code blocks are delivered to the channel coding unit (module). The code blocks are denoted by oir1,oir2,oir3, . . . , oirK
The hybrid ARQ functionality for HS-DSCH is shown in more detail in
As shown in
As also shown in
Nrow=4 for 16QAM and Nrow=2 for QPSK
Ncol=Ndata/Nrow
where Ndata is used as defined in section 4.5.4.3 of 3GPP TS 25.212. The Ndata bits are provided to a physical channel segmentation module (
As further explained in 3GPP TS 25.212 section 5.4, the physical layer HARQ functionality for SISO (and according to the prior art) can be further decomposed as in
Now as explained in section 5.2 of 3GPP TR (Technical Report) 25.876, in case of MIMO processing, i.e. in case of using several different transmit antennae and using per-antenna rate control (PARC) in order to implement FDD high-speed channels such as HS-DSCH, separately encoded (low-rate) data streams—that in combination convey a single high-rate data stream—are transmitted from each antenna with equal power but possibly with different data rates; further, the receiver estimates the channel quality and the information is fed back to the transmitter, which then determines the data rate to use for each antenna.
For HS-DSCH, the basic physical layer structure for PARC is illustrated in
The number of assigned codes may differ from stream to stream (i.e. not all C codes may be used for a low-rate stream), and so the number of OVSF codes for the different (low-rate) streams may be different. If so, the codes for the antenna/low-rate stream with the most codes are assigned first, and then the codes for the other antennas are assigned, using a subset of the already assigned codes, in what is known as “code reuse” because the spreading codes (OVSF codes) are reused among the antennas.
Now referring to
The HARQ bit collector is basically a matrix where bits are written in and later read out in column-by-column order, systematic bits first, followed by redundancy bits (e.g. parity bits), as set out in 3GPP TS 25.212, section 4.5.4.4. In the invention, in the priority multiplexing module the bits that would otherwise be provided to the single HARQ bit collector of the prior art are instead segregated into multiple sets of bits, and each is then provided to a corresponding HARQ bit collector, as shown in
As an example, if a modified HARQ bit collector shown in
Now (and referring to both
According to the invention, Nt,sys,1—the total number of systematic bits to be allocated to the set of Nc2 non-reused codes—should be larger than Nt,sys,2 if possible to achieve better protection for the systematic bits than for the redundancy/parity/non-systematic bits. (Additionally Nt,p1,1 and Nt,p2,1 can each be equal to or close to 0 if desired.) Since the (code) channels using the Nc2 non-reused codes inherently have a higher channel quality (e.g. SINR), allocating bits so that Nt,sys,1 is larger than Nt,sys,2 is equivalent to the rule, provided by the invention, that the systematic bits after turbo encoding are allocated to code channels with a higher channel quality.
The above rule is effective in providing improved protection for the systematic bits at least in cases where the modulation alphabet/symbol constellation is such that all bits have the same SNR after demodulation. In other cases, interleaving the bits evenly in the code domain may result in the best performance.
In case of QPSK modulation, the parity bit allocation on non re-used channel may be 0 but with 16-QAM it is not necessarily desired since half of the bits in 16-QAM modulation have a worse SINR. In that case, a more even distribution might lead into better performance.
It may be assumed that relatively high coding rates and small code differences between the two (or more) bit streams of the invention are normally used. Therefore, at least some systematic bits will often be multiplexed to the code re-used channels as well.
Another possible implementation is to just select the channel segmentation order so that the first columns of the HARQ bit collector matrix are allocated to non-reused codes, since the first columns will contain more systematic bits if the coding rate is such that the number of systematic bits does not fit into the exact number of rows, i.e. if
where └ . . . ┘ represents the integer part of the indicated operation (and is sometimes called the floor function), N1,sys. is the total number of systematic bits (to be conveyed by all the low-rate streams), and Ncol is the number of columns of the HARQ bit collector matrix.
As indicated above, according to the invention, in order to control the bit allocation in the coding chain, the HARQ bit collector (
The channel segmentation unit may be used to prioritize systematic bits as illustrated in
The channel interleaver unit may be used to prioritize systematic bits as illustrated in
Note that during a HARQ retransmission, the invention allows changing the bit allocation, for example by alternating the code channel allocation. In other words, the code channel segmentation could work as in
Besides alternating the code channel allocation, the invention also allows changing the bit allocation by redefining the systematic bit positions (but still per the invention, and so assigning as many as possible to non-reused code channels, and so on). But different bits would end up in different positions. This would eventually average out the differences between the bit SINR values. Although as already mentioned in the HARQ bit collector part, it may be in some situations advantageous to relax the systematic bit allocation strategy and allow parity bits also to be allocated to more favorable positions.
As should be clear from the above, the invention allows for code domain re-ordering during retransmission, i.e. so that the code domain bit allocation may be re-ordered during a HARQ retransmission of a failed packet. In a SISO system, such re-ordering does not affect performance but due to the varying SINR on the code re-used channel, code re-ordering may provide a gain in a MIMO system. One option for implementing code re-ordering is to simply rotate (or otherwise change) the code allocation (i.e. the bit-to-symbol mapping) in the channel multiplexing stage as in
The invention may be used even if a beam-forming type of transformation is used instead of a direct mapping of information streams to transmit antennas if code allocation between the streams may differ. The information streams need not all be targeted to the same user.
The code domain bit ordering strategy for information stream multiplexing type of MIMO scheme provided by the invention is applicable not only for PARC processing, but also for a PURC (Per Unitary Basis Stream User and Rate Control) processing for MIMO (PURC is an extension of PARC, and is explained at e.g. 3GPP WG1 document R1-030354), with or without beam-forming (if a different number of channelization codes on HS-DSCH may possibly be used). The scheme is easier to apply in case of QPSK modulation compared to 16-QAM.
As mentioned, the invention is of use in other than a 3GPP cellular communication system. More specifically, the invention is of use in case of any CDMA system using systematic channel encoding (having separate redundancy/parity bits and information bits), using a MIMO system with flexible code allocation where the encoded transmission is de-multiplexed for several code channels, and the number of code channels is possibly different for the different transmitted information streams, and the information streams may be mapped to the transmit antennas directly (PARC) or through a transformation such as beam-forming (PURC for example, although this scheme also has a more flexible user allocation). The principal objective of the invention is to transmit the systematic bits so that they have the best SNR, and this is done by selecting the code channels so that most of the systematic bits would use the code channel not under code re-use.
Referring now to
Referring now to
It is to be understood that the above-described arrangements are only illustrative of the application of the principles of the present invention. Numerous modifications and alternative arrangements may be devised by those skilled in the art without departing from the scope of the present invention, and the appended claims are intended to cover such modifications and arrangements.