Browse Prior Art Database

©A method for joint transmit-antenna to user association and scheduling in downlink single-user multiple-input multiple-output systems.

IP.com Disclosure Number: IPCOM000179760D
Publication Date: 2009-Feb-25

Publishing Venue

The IP.com Prior Art Database

Abstract

The invention discloses a method for enabling communication between multiple transmitting antennas and multiple users (multiple user receive-antennas) in the downlink communication. The transmitting system can be a system of multiple co-located antennas, or it can be a Distributed Antenna System (DAS) where different subsets of antennas are geographically distributed but work together. In the invention the association between different user-data, the transmit-antennas to transmit each user-data, and the receive-antennas (or users) is flexible and determined by the invention through a joint consideration over many antenna to user assignments. The invention may also additionally determine which users to serve, and which antennas to power on. In the case of a DAS, the system consists of several transmit antennas that are spread geographically over a service area. In terms of the system architecture behind the DAS, the antennas can either be located at two or more base-stations, or may simply be remote antennas spread over an area and connected (e.g. via fiber) to centralized processing units. See Figure 1 to Figure 4 for various examples. A co-located system would simply have all antennas at one geographic location. The invention is concerned with the allocation of transmission resources, e.g. time-frequency slots in an OFDM system or codes in a CDMA system, on each transmit-antenna to the users. In particular it allocates a given resource on each transmit-antenna to a single user, or to a single portion of user-data. One can consider this “user-data” as a unique portion of coded data (transmission symbols) intended for only one user, though in scenarios such as multi-cast a “user” may in fact be a group of users receiving a common piece of data. In the system on a given transmit-antenna, a given transmission resource is not used to send information to two or more users (or user-data), i.e. is not used to send different data to two or more users. This means for example that the system does not consider strategies in which different user-data is precoded or combined, e.g. through linear or other combination, across multiple antennas before transmission within a given transmission resource. Rather the transmit-antenna resource, once the selection and association is made, is dedicated to only one user-data. In this way the process is different from classical MU-MIMO approaches such as those using linear zero forcing precoding, linear precoding, or dirty paper coding. For simplicity, we will consider the assignment of one particular transmission resource over all transmit-antennas, where different antennas can have a different assignment. Therefore the term “transmission resource”, or “transmission resource on each transmit-antenna”, is now synonymous with simply “transmit-antenna”. The system allows any transmit-antenna to be used to transmit data to any user. That is, a user is not necessarily associated to the geographically closest transmit-antenna (or base station supporting this antenna). The user is also not necessarily associated with the transmit-antenna which provides the most favorable channel to that user. Furthermore, as in the case of Multiple Input Multiple Output (MIMO) or Multiple Input Single Output (MISO) systems, a user may receive its (intended) signals from the joint action of different transmit-antennas. Such transmit-antennas, working together, are all assigned to the same user. Base stations and remote locations may have more than one transmit-antenna, and users may have more than one receive-antenna. See for example Figure 5 which illustrates a system were antennas are allocated to different users in a one-to-one or many-to-one fashion. The innovation disclosed in this patent is the ability to jointly determine for a transmission resource across transmit-antennas and users, which transmit-antennas are dedicated to which user. In the method the selection is driven by a priori knowledge (assumed or actual) of the channel state between each transmit-antenna and each user receive-antenna in the system. A channel state may be represented by a single complex number which accounts for any relative path loss, shadowing, short-term phase fluctuations, in the received signal compared to the transmitted signal. If there is one central entity that can direct transmissions on all antennas, as in Figure 3, this entity uses the channel state information to determine the user-data to transmit antenna associations. If antennas are located on two or more base-stations, as in Figure 1 and Figure 2, or two or more central processing units, as in Figure 4, alternatively such entities are given all the same channel state information and can derive the same association by running the same association algorithm remotely and independently using the same state information. The innovation in the method is based on the observation that various interference terms do not change with various association combinations. For example, when an algorithm selects transmit-antennas to be associated with a given user, say “user(1)”, by the constraints in our system these antennas are dedicated only the transmission to “user(1)”. As a result, transmit-antennas not associated with “user(1)”, but which are associated with other users (other user-data), become interference to ``user(1)’’. The interferences does not depend on the precise association of those other transmit-antennas to other users. Note, the association is performed for each given transmission resource. Within a given transmission resource all active transmit-antennas use the same time-frequency slot and/or chip sequence, etc., and as a result interfere with each other. This is why a joint optimization over all users is important and why assigning users simply to the geographically nearest antenna, or the antenna giving the individual best channel to each user, is not necessarily good. It also matters how the transmissions interfere with each other. The selection and association of transmit-antennas to users takes into account both the effective useful signals for each user as well as the generated interferences for each user. From the observation above, once the active set of transmit antennas (and their powers) are known, then if the assignment to “user(1)” is known then also known is the interference to “user(1)” is known. It follows that for a given association across all users and transmit-antennas in a transmission resource, the useful signal to each user, and the interference signal to each user, is therefore known. With this supported transmission rates for various transmission schemes to each user can be determined and can be selected (and used) appropriately for each user. Also note, as in Figure 5, not all transmit antennas have to be used. In fact, it may be beneficial not to power on some antennas since doing so creates interference in additional to useful signals. However, starting with a larger pool of available antennas, though not all will be powered on in the end, creates more diversity to choose from in the channels and is beneficial for boosting up the supported rate. The basic embodiments of the algorithm can be thought of as a (fast, efficient, and/or exhaustive) combinatorial search over all possible users to transmit antennas associations, a search which is simplified by the constraint that each transmission resource on each antenna is used only for one user. That is, assignments of transmit-antennas to users is either one-to-one or many-to-one. For some cases with sufficiently low numbers of users, user antennas, and transmit antennas, a brute-force exhaustive search over all assignment combinations is feasible. Such a search is greatly simplified given the assignment constraint of one-to-one or many-to-one. Searches can also determine, if necessary, which subset of transmit antennas should be powered on in the transmission resource in the case that the system only wants to power on a subset of transmit antennas. For more challenging scenarios, with larger numbers of antennas and users, embodiments of the invention use sub-optimal non-exhaustive searches. These rely on the one-to-one/many-to-one constraint mentioned above that once the powers of transmitting antennas is known, and once a given subset of antennas is selected for a given user, all other powered antennas become interference to this user regardless of how they are assigned to other users. This creates natural tree-search structures in testing various associations. Such trees may be search non-exhaustively using various strategies. In summary, the algorithm provides a method of assigning transmission resource (transmit-antennas) to users through efficient searches that rely on many-to-one or one-to-one assignments. Such assignments have the benefit of defining both the useful signal energy as well as interference energy to each user once an assignment to a user is selected. Furthermore, searchers can be made to satisfy various criteria that are based on such knowledge of useful signal and interference. Examples include making an assignment which performs well in terms of the average rate across users, the minimum rate across users, the proportionally fair rate across users, etc. In addition, the algorithm may be used to schedule users from a pool of users based on their channel characteristics, and/or schedule which transmit-antennas are used (are powered on).

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 8% of the total text.

Page 1 of 25

DoCoMo Communications Laboratories USA, Inc.

3240 Hillview Avenue, Palo Alto, CA 94304, USA

1 TITLE OF THE INVENTION

A method for joint transmit-antenna to user association and scheduling in downlink single-user multiple-input multiple-output systems.

2 INVENTOR(S)

Sean A. Ramprashad

Christine Pépin
Haralabos Papadopoulos
Carl-Erik W. Sundberg

3 FIELD OF THE INVENTION

The present invention relates to Single-User Multiple-Input Multiple-Output (SU-MIMO) wireless systems since each of the scheduled transmissions can be thought of as a SU-MIMO transmission in the presence of scheduled (selected) interference. More particularly, it relates to downlink communication in a Distributed Antenna Systems (DAS) wherein the associations of transmit- antennas to user terminals (or user receive-antennas) is flexible. Furthermore, since the invention schedules multiple concurrent SU-MIMO transmissions, specifying a joint association across all transmit-antennas, users, and user receive-antennas, the invention also relates to Multi-User MIMO (MU-MIMO) and/or Network MIMO (NW-MIMO) systems.

4 SUMMARY OF THE INVENTION

The invention discloses a method for enabling communication between multiple transmitting antennas and multiple users (multiple user receive-antennas) in the downlink communication. The transmitting system can be a system of multiple co-located antennas, or it can be a Distributed Antenna System (DAS) where different subsets of antennas are geographically distributed but work together. In the invention the association between different user-data, the transmit-antennas to transmit each user-data, and the receive-antennas (or users) is flexible and determined by the invention through a joint consideration over many antenna to user assignments. The invention may also additionally determine which users to serve, and which antennas to power on.

In the case of a DAS, the system consists of several transmit antennas that are spread geographically over a service area. In terms of the system architecture behind the DAS, the

1

[This page contains 1 picture or other non-text object]

Page 2 of 25

DoCoMo Communications Laboratories USA, Inc.

3240 Hillview Avenue, Palo Alto, CA 94304, USA

antennas can either be located at two or more base-stations, or may simply be remote antennas spread over an area and connected (e.g. via fiber) to centralized processing units. See Figure 1 to Figure 4 for various examples. A co-located system would simply have all antennas at one geographic location.

The invention is concerned with the allocation of transmission resources, e.g. time-frequency slots in an OFDM system or codes in a CDMA system, on each transmit-antenna to the users. In particular it allocates a given resource on each transmit-antenna to a single user, or to a single portion of user-data. One can consider this "user-data" as a uni...