The present invention relates to a method for communicating between a primary station and a plurality of secondary stations.
This invention is, for example, relevant for telecommunication systems like a mobile telecommunication system. More specifically, this invention is relevant for the UMTS.
In a conventional UMTS system, a PDCCH (Physical Downlink Control Channel) message can use 1, 2, 4 or 8 Channel Control Elements (CCEs or resource elements)—referred to as CCE aggregation levels 1, 2, 4 or 8. A search space is a set of aggregated CCEs (with a certain aggregation level) within which a mobile station (or user equipment (UE) or secondary station) performs blind decoding of all PDCCH payloads possible for that aggregation level. Search spaces are defined per aggregation level; a secondary station in such a system thus can have up to four search spaces. For example, the search space of a UE for aggregation level 1 (referred to as 1-CCE) could consist of the CCEs indexed 3, 4, 5, 6, 7, 8, while its search space for aggregation level 8 could consist of the two resource sets of aggregated CCEs consisting of the CCEs indexed by 1, 2, . . . , 8 and 9, 10, . . . , 16, respectively. in this example, the UE thus performs six blind decodings for 1-CCEs and two blind decodings for 8-CCEs.
In an example, in order to determine the starting point of the search space, mobile stations (or secondary stations, also termed as UEs, for User Equipments in 3GPP parlance) compute a hash function f(UE_ID,s), where UE_ID is the identifier of the UE (different for distinct UEs) and s a time-varying subframe number. It is desirable that different UEs collide (have equal hash value) as infrequently as possible.
The hash function presently proposed within 3GPP is of the form
f(UE_ID,s)=K(UE_ID*16+s)+L modulo M,
where K, L and M are constants, UE ID is the identifier of the UE, and s is the subframe number. It is clear that with this particular hash function f, two UEs that collide for some subframe number collide persistently, i.e., for all subframe numbers.
It is an object of the invention to propose a method for communicating which permits the probability of collisions to be reduced.
Another object of the invention is to provide a method for communicating preventing two UEs from repeatedly colliding.
To this end, according to a first aspect of the invention, a method is proposed for communicating from a primary station to a plurality of secondary stations, comprising the step of at the primary station allocating a resource to the secondary stations over time on the basis of a hash function, wherein the hash function is such that the probability that two secondary stations are allocated common resources in two subframes substantially equals the product of the probability that the two secondary stations are allocated a common resource in the first subframe and the probability that the two secondary stations are allocated a common resource in the second subframe.
As a consequence, the hash functions proposed here aim to reduce the likelihood of persistent collisions. In fact, the hash functions are such that the probability that different UEs collide in two subframes is approximately equal to the probability that two UEs collide in the first of these subframes times the probability that two UEs collide in the second of these subframes. Thus, it is unlikely that two UEs collide repeatedly.
In a specific embodiment of the method, the hash function has the form:
f(x,s)=(h(x)mod g(s))mod M, where x is a parameter of each secondary station, s is the subframe number, h is a function dependent on x, g is a function dependent on s, M is a constant, and mod is the modulo function. In another specific embodiment of the method, h is a constant multiplier.
The present invention also relates to a secondary station comprising means for communicating with a primary, the secondary station further comprising
control means configured to search at least one of a plurality of search spaces, each search space comprising at least one resource set, where at least one resource set might be used to transmit a message to the considered secondary station, wherein the search space of the secondary station is determined on the basis of a hash function, wherein the hash function is such that the probability that two secondary stations are configured to have common resources in the search spaces in both of any two subframes substantially equals the product of the probability that the two secondary stations are configured to have a common resource in the first subframe and the probability that the two secondary stations are configured to have a common resource in the second subframe;
wherein the control means are configured for searching in the configured at least one search space for a control message from the primary station addressed to the considered secondary station, and receiving the control message.
In accordance with still another aspect of the invention, it is proposed a primary station comprising means for communicating with a plurality of secondary stations, the primary station further comprising
allocating means to allocate at least one resource set to a given secondary station into at least one of a plurality of search spaces, each search space comprising at least one resource set, wherein the search space of the given secondary station is determined on the basis of a hash function, wherein the hash function is such that the probability that two secondary stations are configured to have common resources in the search spaces in both of any two subframes substantially equals the product of the probability that the two secondary stations are configured to have a common resource in the first subframe and the probability that the two secondary stations are configured to have a common resource in the second subframe.
These and other aspects of the invention will be apparent from and will be elucidated with reference to the embodiments described hereinafter.
The present invention will now be described in more detail, by way of example, with reference to the accompanying drawings, wherein:
The present invention relates to a method for communicating in a network, like a cellular network. For instance, the network may be a UMTS network as depicted on
Referring to
One of the downlink control channels received by the secondary stations is the PDDCH, where each secondary station has to blindly decode a plurality of sets of CCEs to find which set was allocated to it as set out in the preamble of the description.
In accordance with a first embodiment of the invention, results of various simulations carried out by the inventors are described. With these simulations, it is assumed that 48 CCEs are available. This corresponds to the illustrative exemplary first embodiment of the invention. Various sets of 48 search spaces for the 1-CCEs have been considered; to each user to which a 1-CCE is to be sent, one of these 48 search spaces is assigned at random (the choice corresponds to the outcome of a hash function of that UE that we model as being uniform over the numbers 1, 2, . . . , 48). Each search space consists of six CCEs in this example. The following sets of search spaces have been considered:
In a conventional system, if only sets of 1-CCEs and 8-CCEs are considered, the search space for one secondary station or UE for 8-CCE messages (2 positions 208 are constructed from contiguous groups of CCEs) is depicted on
In accordance with the first embodiment of the invention, the set of available resources 300 comprises search space for one LIE for 8-CCE messages 308, as on
In order to determine the start of the search space of each secondary station, each secondary station uses a hash function. The hash functions disclosed in accordance with this embodiment aim to reduce the likelihood of persistent collisions. In fact, the hash functions are such that the probability that different UEs collide in two subframes is approximately equal to the probability that two UEs collide in the first of this subframes times the probability that two UEs collide in the second of these subframes. Stated differently, collision events in different subframes are approximately independent.
In fact, we describe functions fs(x) with x∈X, s∈{0, 1, . . . , T−1} into {0, 1, . . . , M−1}. The variable x corresponds to the UE_ID in the present situation, and s to the subframe number. The functions have the following properties.
1. For each s∈{0, 1, . . . , T−1}, the function fs attains each element in {0, 1, . . . , M−1} approximately equally often.
2. For all distinct s,t in {0, 1, . . . , T−1}, the number of elements x in X such that fs(x)=i and ft(x)=j is approximately the same for all values of i and j.
We propose to use sets of hash functions of the form
fs(x)=(Ax mod Ms)mod M
where A is a constant number and M0, M1, . . . , MT-1 are different numbers. It is advantageous if M0, M1, . . . , MT-1 are relatively prime to each other and to M.
As a variant of the first embodiment, the following parameters are selected:
T=10, UE ID in X={0, 1, . . . , 224−1}, M=47, and A=1. For the multipliers M0, M1, . . . , M9, we take ten prime numbers close to 212, as depicted in the following table.
To test the “uniformity” of each of the T=10 hash functions, i.e., Property 1 above we counted for i=0, . . . , M−1, the number of elements x∈X for which ft(x)=i. The quotient of the smallest of these numbers and the largest of these numbers are computed. In case of a uniform distribution, this quotient would equal one; we thus wish that the quotient should be approximately one. For our specific choice of M0, M1, . . . , M9, the computed quotients range from 09885 to 09906,
To test the independence of the hash functions fs and ft, i.e. Property 2 above, we computed for all pairs (i,j) the number elements x∈X for which fs(x)=i and ft(x)=j,
Next, we computed the quotient of the smallest of these M2 number and the largest of these M2 numbers. Ideally, we would like this quotient to be equal to one. For our specific choice of M0, M1, . . . , M9, the computed quotients range from 0.9752 to 0.9808.
We can conclude that in the embodiment, the hash functions are approximately uniform and approximately independent.
In the envisioned application, the values of T and the range X is fixed while M may vary. For implementation reasons, it is advantageous that M0, M1, . . . , MT−-1 do not depend on M. If we change M to 24, the computed quotients for uniformity range from 0.9941 to 0.9952; the computed quotients for testing independence range from 0.9779 to 0.9889. So also for this case, the proposed hash functions are approximately uniform and approximately independent. If we change M to 120, the computed quotients for uniformity range from 0.9706 to 0.9762; those for testing independence range from 0.9330 to 0.9474.
The invention may be applicable to mobile telecommunication systems like UMTS LTE and UMTS LTE-Advanced, but also in some variants to any communication system having allocation of resources to be done dynamically or at least semi persistently.
In the present specification and claims the word “a” or “an” preceding an element does not exclude the presence of a plurality of such elements. Further, the word “comprising” does not exclude the presence of other elements or steps than those listed.
The inclusion of reference signs in parentheses in the claims is intended to aid understanding and is not intended to be limiting.
From reading the present disclosure, other modifications will be apparent to persons skilled in the art. Such modifications may involve other features which are already known in the art of radio communication.
Number | Date | Country | Kind |
---|---|---|---|
08305070 | Mar 2008 | EP | regional |
Number | Name | Date | Kind |
---|---|---|---|
8385960 | Skoric et al. | Feb 2013 | B2 |
20030086384 | Kwon et al. | May 2003 | A1 |
Entry |
---|
Ericsson, “Summary of Email Discussion on DL Control Signaling”, 2GPP Draft, TSG-RAN WG1#52, R1-080869, Feb. 2008, 8 Page Document. |
Motorola, “Separch Space Definition for L1/L2 Control Channels”, 3GPP TSG RAN1#51, R1-073373, Aug. 2007 Retrieved Fromthe Internet: http//www.3gpp.org/ftp/tsg—ran/WG1—RL1/TSGR1—50/Docs/R1-073373.zip; 11 Page Document, Retrieved Jan. 29, 2009. |
Ericsson, “PDCCH Blind Decoding-Outcome of Offline Discussions”, R1-081101, Feb. 2008, Retrieved Fromt He Internet: http//ftp.3gpp.org/tsg—ran/WG1—RL1/TSGR1—52/Docs/?C+N;0=d>, 7 Page Document. |
Philips, “Analysis of Search Space Design for PDCCH Blind Decoding”, 3GPP TSG RAN WG1#52bis, R1-081504, Mar. 26, 2008, 5 Page Document. |
Number | Date | Country | |
---|---|---|---|
20130155985 A1 | Jun 2013 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 12933911 | US | |
Child | 13766850 | US |