Search space for ePDCCH control information in an OFDM-based mobile communication system

Abstract

The present invention relates to a method for receiving control information within a subframe of a multi-carrier communication system supporting carrier aggregation, the method comprising the following steps performed at a receiving node: performing a blind detection for the control information within a search space by means of a first search pattern, wherein the first search pattern is one of a plurality of search patterns, each of the plurality of search patterns comprising a plurality of candidates distributed on any of a plurality of aggregation levels, and wherein the plurality of search patterns further comprises a second search pattern whose candidates are non-overlapping the candidates of the first search pattern on the same aggregation levels.

Claims

1. An integrated circuit configured to operate a transmitting apparatus, the integrated circuit comprising: control circuitry, which, in operation, maps downlink control information to at least one of a first search space or to a second search space, the first search space including a first set of PDCCH candidates corresponding to a plurality of aggregation levels, the second search space including a second set of PDCCH candidates corresponding to the plurality of aggregation levels and an aggregation level higher than any one of the plurality of aggregation levels, wherein for each of the plurality of aggregation levels a resource for each PDCCH candidate of the first set of PDCCH candidates does not overlap with a resource for any PDCCH candidate of the second set of PDCCH candidates; and transmitting circuitry, which is coupled to the control circuitry and which, in operation, transmits the mapped downlink control information.

2. The integrated circuit according to claim 1, wherein the first set of PDCCH candidates included in the first search space are allocated for localized transmission; and the second set of PDCCH candidates included in the second search space are allocated for distributed transmission.

3. The integrated circuit according to claim 1, wherein resources for the first set of PDCCH candidates do not overlap with resources for the second set of PDCCH candidates.

4. The integrated circuit according to claim 1, wherein the first set of PDCCH candidates and the second set of PDCCH candidates are allocated for distributed transmission.

5. The integrated circuit according to claim 1, wherein at least one of the first search space and the second search space defines a smaller number of PDCCH candidates as the aggregation level gets higher.

6. The integrated circuit according to claim 1, wherein both of the first search space and the second search space are UE-specific search spaces.

Description

(1) The above and other objects and features of the present invention will become more apparent from the following description and preferred embodiments given in conjunction with the accompanying drawings in which:

(2) FIG. 1 is a schematic drawing showing an exemplary downlink component carrier of one of two downlink slots of a sub-frame defined for 3GPP LTE release 8;

(3) FIG. 2 is a schematic drawing illustrating the structure of a non-MBSFN sub-frames and a physical resource block pair thereof defined for 3GPP LTE release 8 and 3GPP LTE-a release 10;

(4) FIG. 3 is a schematic drawing illustrating a structure of MBSFN sub-frames and a physical resource block pair thereof defined for 3GPP LTE Release 8 and 3GPP LTE-A Release 10;

(5) FIG. 4 is a schematic drawing of an exemplary network configuration including a donor eNodeB, a relay node, and two user equipments;

(6) FIG. 5 schematically illustrates possible combination of UE scenarios in accordance with an embodiment of the present invention;

(7) FIG. 6 schematically illustrates search patterns for two UE from the same UE scenario;

(8) FIGS. 7-10 schematically illustrates search patterns in accordance with embodiments of the present invention;

(9) FIG. 11 schematically illustrates a search pattern configuration in accordance with an embodiment of the present invention;

(10) FIG. 12 schematically illustrates a further pattern design in accordance with an embodiment of the present invention;

(11) FIG. 13 schematically illustrates a search pattern configuration in accordance with an embodiment of the present invention;

(12) FIG. 14 schematically illustrates a search pattern configuration in accordance with an embodiment of the present invention; and

(13) FIG. 15 schematically illustrates further search patterns in accordance with an embodiment of the present invention.

(14) Thanks to the search space design of the present invention it is possible to avoid the complexity of full flexibility, while providing sufficient choices for different scenarios with limited number of blind decoding trials.

(15) In the following, it is assumed that legacy PDCCH concept is reused, i.e. one ePDCCH is aggregation of {1, 2, 4, 8} eCCEs. It is also assumed that one PRB pair is divided into four eCCEs.

(16) With reference to FIG. 5, a number of different scenarios can be defined in the following manner. According to UE position, there are mainly three scenarios: 1. scenario 5101 comprising cell-center UEs, which can be configured, for instance, with more lower aggregation level candidates; 2. scenario 5103 comprising cell-middle UEs can be configured with some higher aggregation level candidates and some lower aggregation level candidates 3. scenario 5102 comprising cell-edge UEs can be configured with more higher aggregation level candidates; and

(17) At the same time, according to UE feedback, there are mainly three scenarios: i. scenario 5201 comprising UE with more accurate feedback, for instance moving at low speed, preferably using localized candidates; ii. scenario 5202 comprising UE with less accurate feedback, for instance moving at high speed, preferably using distributed candidates; and iii. scenario 5203 comprising UE with roughly accurate feedback, preferably using both localized and distributed candidates.

(18) Accordingly, in order to provide a targeted search space for all possible combinations of scenarios 5101-5103 and scenarios 5201-5203, nine possible search patterns have to be defined. However, associating one search pattern to each possible combination may cause blocking. Moreover, such an approach makes it difficult to pack different DCI messages within the same PRB pair.

(19) For instance, with reference to FIG. 6, it can be seen how UE1 and UE2, both being, for instance, cell-middle UEs with less accurate feedback, would have the same search pattern. Accordingly, this makes it difficult to multiplex search spaces from different UE within the same PRB pair. In fact, in such a situation, only spatial multiplexing is possible by, as indicated in the figure, allocating UE1 to AP8 and UE2 to AP7. However, if there are many such kinds of UEs in the system, blocking among search space becomes increasingly critical.

(20) This can be improved by providing a plurality of search patterns having a certain number of candidates for one or more aggregation levels in such a manner to avoid overlapping of search patterns on the same aggregation level for at least two patterns.

(21) More specifically, FIG. 7 schematically illustrates two patterns, pattern 0 and pattern 1, in accordance with an embodiment of the present invention.

(22) In particular, in FIG. 7 the horizontal axis represents the VRB index; the vertical axis represents the AP value while the remaining axis represents the aggregation level. The two patterns 0 and 1 comprise each a plurality of candidates arranged on any of aggregation levels 1, 2, 4 and/or 8. As can be seen, pattern 0 has candidates on aggregation level 1 and aggregation level 2. Similarly, pattern 1 also has candidates on aggregation level 1 and on aggregation level 2. Additionally, the two patterns are designed so that they are non-overlapping. In particular, the mapping of the candidates on aggregation level 1 of pattern 0 does not overlap with candidates on aggregation level 1 of pattern 1. Similarly the mapping of the candidates on aggregation level 2 of pattern 0 does not overlap with candidates on aggregation level 2 of pattern 1.

(23) Alternatively, or in addition, patterns 0 and 1 are designed such that the same aggregation levels and the corresponding number of candidates are present. Further alternatively, or in addition, the mapping of candidates to eCCEs is complementary in both sides for the respective aggregation levels, that is, eCCEs for aggregation level 1 in pattern 0 are not used for aggregation level 1 in pattern 1, and similarly for aggregation level 2.

(24) By defining pattern 0 and pattern 1 in such a manner, packing, i.e. multiplexing, of different DCI messages in the same PRB is achieved since the patterns do not overlap. In particular, DL and UL assignments to the same UE are possible to be transmitted in the same PRB pair. Moreover, since both pattern 0 and pattern 1 define the same number of candidates on the same aggregation levels, they can be applied to different UEs in the same scenario, for instance they could be applied, respectively, to UE1 and UE2 of FIG. 6, without overlapping. This provides more flexibility as the number of possible active UEs can be increased without blocking arising on the channel.

(25) Alternatively, or in addition, FIG. 8 schematically illustrates a further criterion for the definition of a further search pattern, in accordance with an embodiment of the present invention.

(26) In particular, pattern 0 of FIG. 8 corresponds to pattern 0 already defined in FIG. 7. Pattern 3, illustrated in FIG. 8 is constructed so as to provide higher aggregation level candidates, compared to pattern 0, while still providing non-overlapping candidates on aggregation level 2 with respect to pattern 0. This provides the possibility of employing at the same time, both pattern 0 and pattern 3.

(27) Moreover, this allows DCI messages from UEs configured with higher aggregation level candidates to be multiplexed with DCI messages from UEs configured with more lower aggregation level candidates. Even for the same UE, candidates of aggregation levels 1, 2 and 4 can be configured so that US search space does not need to be reconfigured even if UE scenario changed.

(28) Additionally, this design is advantageous since it allows different patterns to have candidates on different aggregation level. For instance, cell-center UEs can be associated with patterns having lower aggregation level candidates, such as pattern 0. At the same time, cell-edge UEs can be associated with patterns having higher aggregation level candidates, such as pattern 3. In this manner, with a limited number of blond decoding trials, different UEs can be configured with different numbers of lower aggregation level candidates and higher aggregation level candidates.

(29) Alternatively, or in addition, FIG. 9 schematically illustrates a further criterion for the definition of a further search pattern, in accordance with an embodiment of the present invention.

(30) In particular, FIG. 9 illustrates a pattern 4 in which only candidates from the largest aggregation level are used. Such an approach provides the advantage that spatial and/or frequency diversity can be obtained, at least for the largest aggregation level, as a fallback mode. Moreover, another benefit is that since aggregation level 8 candidates can easily block candidates of other aggregation levels, pattern 4 can always be configured on another antenna port to avoid blocking of candidates of other aggregation levels.

(31) Although in the above embodiments only five search patterns have been defined, the present invention is not limited thereto and the number of patterns can be increased, by constructing other patterns in accordance with the rules given above, or reduced.

(32) FIG. 10 schematically illustrates the combination of five potential search patterns in accordance with an embodiment of the present invention.

(33) As can be seen pattern 0 and 1, as well as pattern 2 and 3 offer complementary candidates. This is turn allows packing of different DCI messages in the same PRB(s). Moreover, pattern 0 and 1 offer candidates for lower aggregation level, while Pattern 2 and 3 offer candidates mainly for higher aggregation levels. This is beneficial since with a limited number of blind decoding trials, different UEs can be configured with different number of lower aggregation level candidates and higher aggregation level candidates. Additionally, pattern 4 offers candidates for AL 8 so that spatial and/or frequency diversity can be obtained at least for the largest aggregation level as fallback mode. Moreover, it can be seen that the patterns are such that candidates are not overlapping on the same aggregation level.

(34) When employing the search patterns as described above, it is possible to define a search space by configuring the patterns with the following parameters: pattern ID, such as pattern 0, 1, 2 and/or 3, as defined above; and/or antenna port, determining which DM-RS port is used to demodulate the ePDCCH; and/or RB set, determining on which RBs the eCCEs should be detected; and/or diversity configuration, determining whether e.g. LVRB, DVRB, SFBC is used for mapping on PRB.

(35) In particular, the antenna port can be used to define the DM-RS port the pattern is mapped to, thereby defining the spatial domain. The advantage of such parameter is that it offers spatial scheduling gain thereby allowing more candidates in the spatial domain and that it offers the possibility to more candidates to avoid blocking. The RB set can be used to determine the set of RBs the pattern is mapped to thereby defining the frequency domain. The advantage of such parameter is that offers frequency scheduling gain thereby allowing more candidates in the frequency domain and that it offers the possibility to more candidates to avoid blocking as well. Finally, the diversity configuration can be used to determine whether, for instance, LVRB, DVRB, SFBC is used for mapping on PRB. The advantage of such parameter is that it offers spatial and/or frequency diversity when the channel is not known such as, for instance, when frequency/spatial selective scheduling is not feasible.

(36) An exemplary configuration is schematically illustrated in FIG. 11, in accordance with an embodiment of the present invention.

(37) In particular, the configuration comprises: a UE1 being a cell-middle UE with less accurate feedback, as in the case of FIG. 6, and configured with pattern 3 on AP8, in distributed mode, and Pattern 4 on AP7 in distributed mode; and a UE2 being a cell-middle UE with less accurate feedback, as in the case of FIG. 6, and configured with pattern 2 on APB, in distributed mode, and Pattern 4 on AP7 in distributed mode.

(38) Accordingly, UE1 and UE2 being in similar conditions can use complementary patterns so as to achieve the same performances. Thanks to such configuration, AL2 and AL4 candidates of UE1 and UE2 search spaces can be multiplexed within one PRB pair since patterns 2 and 3 are complementary. Accordingly, this allows packing, in other words, multiplexing, of different DCI messages in the same PRB(s). At the same time, there is no blocking of AL2 and AL4 candidates from UE1 and UE2. Additionally, pattern 4 contains two AL8 candidates; accordingly there is no blocking of AL8 candidates from UE1 and UE2. Moreover, since AL8 candidates are configured on AP7, there is no blocking between AL8 candidates and AL2/AL4 candidates.

(39) Accordingly, the above described configuration based on the above described patterns provides sufficient flexibility of search space configuration for different UE scenarios with limited complexity compared with full flexibility

(40) In particular, in full flexibility the number of candidates is equal to
[N.sub.PRB.Math.4.Math.12.Math.2].sup.40=560bits
where NPRB is the number of PRBs within the whole bandwidth. For instance, NPRB is equal to 100 for 20 MHz. 4 is the number of APs, 12 is the number of candidates within one PRB pair, and 2 is the number of diversity choices.

(41) On the other hand, with the present invention, the number of candidates is equal to

(42) $\left(\begin{array}{c}{N}_{PRB}\\ 4\end{array}\right).Math.4.Math.12.Math.2=28\mathrm{bits}$
per each pattern. If a maximum of 3 or 4 patterns, for instance, is configured for one UE, then 84-112 bits are required. Accordingly, the present invention uses a very reduced signaling overhead, when compared with full flexibility approach.

(43) Moreover, the invention supports fallback mode by obtaining frequency and/or spatial diversity at least for the largest aggregation level. Additionally, it support frequency ICIC by allowing packing of multiple DCI messages in the same PRB(s). Moreover, it avoids blocking of candidates by different DCI messages both from the same or different UE. Finally, it provides a SS framework, which allows operating various different network policies to schedule and/or configure ePDCCHs, for instance depending on operating preferences and/or deployment scenarios.

(44) While in the above described embodiment a search space is defined by configuring a set of patterns having as parameters the antenna port, and/or the RB set and/or the diversity configuration, the present invention is not limited thereto.

(45) Alternatively, or in addition, an applicable set of subframes can be added to the search space configuration, thereby providing a time domain diversity as well. In particular on high-interference subframes, and/or when common search space needs to be monitored, a larger number of higher aggregation level candidates, that is, patterns, can be configured, while on low-interference subframes, a larger number of lower aggregation level candidates, that is, patterns, can be configured, so as to save resources.

(46) As an example, the set of subframes can be tied to the subset definitions for CSI reporting. Alternatively, or in addition, the set of subframe can be tied to low-power ABS subframes and non-lower-power ABS subframes.

(47) FIG. 12 illustrates a further pattern design separated by aggregation levels in accordance with an embodiment of the present invention.

(48) In this embodiment, the patterns are designed according to aggregation levels. In particular, each pattern contains candidates of one aggregation level. Moreover, for aggregation level 1, 2 and 4, there are two patterns that are complementary to each other. Additionally, the figure illustrates, to the right of each pattern, the corresponding number of candidates, such as Nc=8 for patterns 0 and 1.

(49) This solution provides the benefit of a more flexible combination and configuration of the patterns.

(50) FIGS. 13 and 14 schematically illustrate search pattern configurations in accordance with further embodiments of the present invention.

(51) In particular, in FIG. 13, a cell-center UE with less feedback is configured with pattern 0 and 1 from FIG. 10, with distributed transmission. In particular, the top part of FIG. 13 illustrates the two patterns: SS1: Pattern 0, AP 7, VRB set 0, DVRB SS2: Pattern 1, AP 8, VRB set 0, DVRB

(52) while the bottom part of FIG. 13 illustrates the resulting configuration.

(53) Additionally, in FIG. 14, a cell-center UE with less feedback is configured with pattern 2, 3 and 4 from FIG. 10, with distributed transmission. In particular, the top part of FIG. 14 illustrates the three patterns: SS1: Pattern 2, AP 7, VRB set 0, DVRB SS2: Pattern 3, AP 7, VRB set 0, DVRB SS3: Pattern 4, AP 8, VRB set 0, DVRB

(54) while the bottom part of FIG. 14 illustrates the resulting configuration.

(55) Moreover, FIG. 15 schematically illustrates further search patterns in accordance with an embodiment of the present invention.

(56) In particular, in FIG. 15, all the candidates within one pattern do not overlap with each other, so that there is no blocking of candidates within one pattern. Alternatively, or in addition, pattern 0 and 1, as well as 0 and 3, have complementary candidates. Further alternatively, or in addition, aggregation level 2 on pattern 3 and aggregation level 1 on pattern 0 offer complementary candidates.