This application relates to wireless communication, and more to particularly to the implementation of carrier aggregation in wireless communication.
The famous Shannon's law for communication establishes a linear proportionality between available channel bandwidth and the amount of data that can be transmitted through the corresponding channel. As determined by this law, higher data rates require more bandwidth at a given signal-to-noise ratio (SNR) as opposed to lower data rate communications at the same SNR. But a given amount of bandwidth has a relative amount of worth: signal attenuation is markedly higher as frequency increases. Thus, it is better to have bandwidth in the regulated spectrums such as at 700 MHz as opposed to having the same amount of bandwidth in the unregulated higher frequency bands such as at 2.4 GHz.
Despite the scarceness of desirable spectrums for wireless communications, the requirement for additional bandwidth is ever increasing. Indeed, regardless of the particular frequency for wireless communication, the need for bandwidth is non-negotiable if one wants to achieve higher data rates. Modern 4G telecommunication protocols such as Long Term Evolution-Advanced (LTE-A) are proposing 1 Gps (one billion bits per second) downlink data rates or even higher. But it is difficult to achieve such a data rate in the limited communication bandwidths that are available to a wireless carrier, particularly in the desirable “beachfront” spectrums such as 700 MHz. For example, the current generation of LTE uses orthogonal subcarriers spread across a channel bandwidth that may range from 1.4 MHz to a maximum of 20 MHz. The subcarriers are separated by 15 KHz such that the maximum symbol rate for each subcarrier is thus 15,000 symbols/second. The number of bits per symbol depends upon the modulation scheme—LTE supports a maximum of 64 bits per symbol using 64QAM. Thus, the 20 MHz channel of LTE supports a raw data rate of 108 Mbps. The actual data rate will depend upon coding overhead and other variables. One can thus appreciate that if LTE-A is to achieve a 1 Gps data rate, the channel bandwidth must be increased by multiples of the LTE 20 MHz channel But note that backward compatibility with conventional LTE should be maintained. Thus, carrier aggregation in LTE-A involves the use of multiple 20 MHz channels. To a conventional LTE handset (which may be designated as user equipment (UE)), each 20 MHz channel operates as a conventional LTE channel. But to an LTE-A UE, data can be received across multiple combinations of such channels. Since each LTE channel corresponds to an LTE carrier, the LTE carrier becomes a component carrier for an LTE-A UE. Carrier aggregation thus preserves precious bandwidth resources for conventional lower-data-rate communication yet achieves greater bandwidth resources for high-data-rate communication.
One of the main technical challenges for implementing carrier aggregation in LTE-Advanced systems is the backward compatibility requirement with the current LTE systems. The additional bandwidth provided by carrier aggregation provides an opportunity for frequency diversity. But because of the complications raised by the need for backwards compatibility, existing carrier aggregation schemes do not exploit frequency diversity. Instead, conventional carrier aggregations schemes enjoy frequency diversity only within each component carrier—for example, a conventional uplink LTE channel is interleaved. Accordingly, there is a need in the art for improved carrier aggregation schemes that exploit the opportunity for frequency diversity across the component carriers rather than just within each component carrier.
In accordance with an aspect of the disclosure, a method is provided that includes the acts of providing a plurality of transport blocks, each transport block corresponding to a component carrier (CC); in a baseband processor, channel coding each transport block into a corresponding channel-coded data signal; in the baseband processor, bit-combining the channel-coded data signals into a bit-combined data signal; and in the baseband processor, interleaving the bit-combined data signal to produce an interleaved plurality of code words.
In accordance with another aspect of the disclosure, a downlink method is provided that includes the acts of determining whether a plurality of component carriers are being interleaved; if a plurality of component carriers are being interleaved, bit-combining a plurality of channel-coded data signals to form a bit-combined data signal; writing the bit-combined data signal into an interleaver matrix stored within a memory, wherein the interleaver matrix is arranged into a plurality of sub-matrices corresponding to the plurality of component carriers; reading from each sub-matrix to retrieve a corresponding output data signal; and modulating each component carrier according to the corresponding output data signal.
In accordance with yet another aspect of the disclosure, a wireless device, is provided that includes a memory; a baseband processor configured to channel code a plurality of transport blocks into a corresponding plurality of channel-coded data signals, bit-combine the channel-coded data signals into a bit-combined data signal, write the bit-combined data signal into an interleaver matrix stored within the memory, and to read from the interleaver matrix to produce an interleaved data signal; and a radio-frequency integrated circuit (RFIC) configured to modulate an RF carrier signal according to the interleaved data signal.
Frequency diversity carrier aggregation is described herein with regard to a Long Term Evolution Advanced (LTE-A) implementation. However, it will be appreciated that the principles of the disclosed carrier aggregation are readily applicable to other wireless communication protocols such as WiMax. The carrier aggregation of the present application is denoted as frequency diversity carrier aggregation in that frequency diversity across the aggregated component carriers is advantageously achieved yet backwards compatibility with conventional LTE (no carrier aggregation) is maintained. This compatibility is best understood with regard to the shared channel, which is used to transmit both data and some control information.
The shared channel data and control information passes from the MAC layer in LTE systems to the physical (PHY) layer through transport channels, which form the interface between the MAC and PHY layers. The uplink and downlink transport channels process data in transport blocks, which are groups of resource blocks sharing a common modulation and coding implementation. In addition to a shared transport channel in both the uplink and downlink, there are other types of transport channels such as a broadcast channel and a random access channel. But since the focus of carrier aggregation is to increase data rate, only the data-carrying shared channels are discussed herein. To illustrate the difficulties of maintaining backward compatibility, the LTE conventional processing of the downlink and uplink shared channels will be discussed and contrasted with the carrier aggregation processing for these channels. The uplink shared transport channel will be discussed first followed by the downlink shared transport channel
Turning now to the drawings, the transport channel processing for a conventional LTE uplink shared channel (UL-SCH) is illustrated in
The control data for the transport block arrives at channel coding module 110 in three forms: channel quality information (CQI), rank indication (RI), and hybrid automatic repeat request acknowledgment (HARQ-ACK). The corresponding channel coded signals are represented by vectors q0ACK, q1ACK, . . . , qQ′
Prior to interleaving, the CQI encoded sequence (represented by the vector q0RI, q1RI, q2RI, . . . qQ′
Channel interleaver 120 interleaves such that HARQ-ACK indications are present on both slots in a subframe. The number of modulation symbols in each subframe is given by H″=H′+Q′RI. As defined by LTE Release 9, an output bit sequence from interleaver 120 represented by h0, h1, h2, . . . , hH+Q
y
r×C
+
c
= qiRI
The variable Column Set is given in Table 1 and indexed left to right from 0 to 3.
Having thus written the RI data to the output matrix (if there is such data to be written), interleaver 120 may then process the multiplexed data and CQI information in a step 215 as follows: interleaver 120 writes the input vector sequence, for k=0, 1, . . . H′−1, into the (Rmux×Cmux) matrix by sets of Qm rows starting with the vector y0 in column 0 and row 0 to (Qmux−1) and skipping the matrix entries that are already occupied:
The pseudocode is as follows:
The HARQ-ACK information (if present) is written last to the output matrix by interleaver 120. Thus, if HARQ-ACK information is to be transmitted in the current subframe, a step 220 tests for whether the RI information and the multiplexed data and CQI information has been already interleaved. Only after all the other types of input sequences have been interleaved does interleaver 120 finally interleave the HARQ-ACK information in a step 225 as follows: the vector sequence q0ACK, q1ACK, q2ACK, . . . , qQ′
y
r×C
+c
= qiACK
The Column Set is given in Table 2 and indexed left to right from 0 to 3. The output of interleaver 120 is the bit sequence read out column-by-column from the (Rmux×Cmux) matrix constructed as just discussed. The bits after channel interleaving are denoted by h0, h1, h2, . . . , h+Q
Having thus constructed the output matrix, which can be stored in memory as discussed above, interleaver 120 may then read out the output matrix column-by-column in a step 230 to finish the interleaving process. The end result of this processing of a transport block is typically denoted as an LTE codeword. The conventional LTE downlink shared channel will now be discussed.
The transport channel processing for a conventional LTE downlink shared channel (DL-SCH) is shown in
However, all the mechanisms discussed above with regard to
To exploit the enhanced frequency diversity opportunity presented by carrier aggregation (CA), an interleaver functioning across the different CCs is disclosed herein for CA systems. In this fashion, frequency diversity is exploited in carrier aggregation by interleaving bits across component cartiers. In general, backward compatibility with conventional LTE is a significant problem. However, backward compatibility is advantageously achieved by the disclosed frequency diversity technique as discussed further herein. In the downlink shared channel, the disclosed CA channel interleaver is added over the CCs, while for the uplink shared channel the proposed interleaver just takes place of the conventional LTE channel interleaver. The CA channel interleaver functions as a conventional LTE channel interleaver when there is only one CC. The CA channel interleaver exploits enhanced frequency and time diversity with the advantage of easy implementation.
To better illustrate the disclosed CA channel interleaver, the following discussion assumes that there are N CCs, where N is some positive integer. As shown in
Bit combiners 422 and 423 perform analogous bit combinations on the N channel-coded RI input signals and the N channel-coded HARQ-ACK input streams for the N transport blocks being interleaved. Bit combiner 422 thus produces a bit-combined RI output signal designated as [q′0RI, q′1RI, q′2RI, . . . , qNQ′
The second stage for CA channel interleaver 420 is a channel interleaver 425 that interleaves the three bit-combined output signals produced in the bit-combining first stage. The number of modulation symbols in each subframe is given by H″=N (H′+Q′RI). Channel interleaver 425 is configured to derive its output bit sequence as follows: Interleaver 425 writes to an output matrix that may be stored in a memory or buffer as analogously described above with regard to conventional LTE processing. The number of columns for this output matrix is given by Cmux=NsymbPUSCH. The columns of the matrix are numbered 0, 1, 2, . . . , Cmux−1 from left to right as also previously discussed. However, the number of rows is given by Rmux=(H″·Qm)/Cmux, which is N times of the number of rows in LTE UL. Each continuous block of Rmux/N rows in the output matrix may be considered to form a sub-matrix that corresponds to one CC. There are thus N sub-matrices in the output matrix corresponding to the N CCs.
RI information is processed in a step 520 by being segmented into N equal subsequences. For example, if the input to step 510 is considered to form an input signal [a1, a2, . . . , an], then the output from step 520 forms the N subsequences [a1, a2, . . . , an/N], . . . , [an-n/N+1, an-n/N+2, . . . , an]. Each subsequence corresponds to a CC transport block. Each subsequence is interleaved into the corresponding carrier component sub-matrix in a step 525 following the way discussed above with regard to step 210 of
With RI information interleaving completed, the data/CQI information may interleaved in a step 530 by writing the input vector sequence, for k=0, 1, . . . , NH′−1 into the (Rmux×Cmux) output matrix by sets of Qm rows starting with the vector y0 in column 0 and rows 0 to (Qm−1) and skipping the matrix entries that are already occupied by RI information as:
where R′mux=Rmux/Qmux.
The HARQ-ACK information is written into the output matrix only after the RI information and the data/CQI information has been processed. Thus, a step 535 delays the interleaving of the HARQ-ACK information accordingly. Once step 535 determines that the RI information and the data/CQI information has been processed, the HARQ-ACK information is segmented in a step 540 in same way as discussed with regard to step 525. Each resulting subsequence corresponds to a carrier component and is interleaved in a step 545 into the corresponding CC sub-matrix as discussed with regard to step 225 of
With the output matrix thus completed, the component carrier data may be read from the corresponding sub-matrix column-by-column in a final step 550. The result would be N output code words for the N component carriers. It can readily be seen that if N=1, the CA channel interleaver 420 performs exactly the same as the conventional 120 channel interleaver discussed with regard to
As shown in
The resulting bit-combined output from combiner 630 is received by a carrier aggregation channel interleaver 640.
Each carrier component is read from its sub-matrix column-by-column in a step 720 to complete the downlink processing. Each sub-matrix thus corresponds to a component carrier code word. One can observe from
The above carrier aggregation process may be entirely implemented at baseband and is thus readily implemented in a baseband processor.
Baseband processor 815 may be programmable such that it implements the downlink or uplink modules discussed above using software implemented on a microprocessor or through programmed logic resources within an FPGA. Alternatively, baseband processor 815 may be a dedicated ASIC. Regardless of how the baseband processing is implemented, it will advantageously interleave the downlink or uplink shared channel across the component carriers to exploit frequency diversity as discussed herein.
Embodiments described above illustrate but do not limit the disclosure. It should also be understood that numerous modifications and variations are possible in accordance with the principles of the present disclosure. For example, although the frequency diversity exploitation discussed above was regard to an LTE enhancement, it will be appreciated that the same technique can be readily applied to other high speed wireless protocols such as WiMax. Accordingly, the scope of the disclosure is defined only by the following claims.
This application claims the benefit of U.S. Provisional Application No. 61/318,696, filed Mar. 29, 2010.
Number | Date | Country | |
---|---|---|---|
61318696 | Mar 2010 | US |