METHOD FOR PERFORMING MULTIPLE ACCESS IN WIRELESS OFDM CELLULAR SYSTEMS OVER MULTIPATH WIRELESS CHANNELS CONSIDERING BOTH SPACE AND FREQUENCY DOMAINS, BASE STATION AND COMPUTER PROGRAMS THEREOF

摘要

Methods, a base station and computer programs for performing multiple access in wireless OFDM cellular systems over multipath wireless channels considering both space and frequency domains,;The base station 100 comprises a large number of antennas 103 in the form of a two-dimensional array and transmits a given number of signals to a number of users 106, so that each one receives its intended signal (or set of intended signals) without interference to/from the other users 106. The base station 100 includes a scheduler in space-time-frequency dimensions 101 as well as an orthogonal space-frequency processing technique 102 for addressing the users 106. The cellular scenario is assumed where the wireless channel 105 presents significant multipath, thereby resulting in multiple signal components being received by each user 106. Orthogonality of the beams is achieved upon transmission by properly discretizing the spatial domain, while inter-user interference is avoided by exciting only those beams that result in multipath components received by one single user 106.

主权项

1. A method for performing multiple access in wireless OFDM cellular systems over multipath wireless channels considering both space and frequency domains, wherein at least one base station equipped with a large number of antennas according to a two-dimensional rectangular array has complete knowledge of the channel characteristics between it and each one of a number M′ of cell users, said rectangular array comprising N1 antenna elements along one axis with a regular spacing dx, and N2 antenna elements along a perpendicular axis with a regular spacing dy, and each one of said users being characterized by angles (θ,φ) in a spherical coordinate system discretized through a grid spacing (Δu,Δv) in the (u, v) domain to achieve orthogonality by letting u=sin(θ)cos(φ), v=sin(θ) sin(φ), dx=λ/(N1Δu), dy=λ/(N2Δv), uk=k·Δu and vl=l·Δv, for k=0, 1, . . . , N1−1, l=0, 1, . . . , N2−1,the method being characterized in that the at least one base station comprises performing following steps:
obtaining, from known training or pilot signals transmitted by each one of the users, signals in the (u, v) domain by means of:SR[k,l,f]=1NcN1N2∑t=0Nc-1∑n=0N1-1∑m=0N2-1AR[n,m,t]exp(-j2πNcft)exp(j2πN1nk)exp(j2πN2ml) where AR[n,m,t] denotes the time-domain signal received by antenna element (n, m), Nc denotes the number of subcarriers in the frequency domain, and SR[k,l,f] denotes the frequency components of the received beam in (k, l) direction; estimating angular channel coefficients corresponding to each one of the users by means of using the following equation:
SR[k,l,f]=Hi[k,l,f]·Piloti[f]+N, where Piloti[f] denotes the a-priori known pilot/training signals transmitted by user i, Hi[k,l,f] is the channel frequency response associated with user i in the direction (k, l), and N is a complex additive noise component;
estimating an angular profile Φi characterizing each one of the users, said angular profile Φi comprising a set of directions, or beams, (k, l) in the (u, v) domain for which the estimated angular channel coefficients Ĥi[k,l,f] are not zero:
Φi={(kj,lj),∀j=0, . . . ,N1N2−1 such that Ĥi[k,l,f]≠0}; discarding, from said estimated angular profiles Φi, those beams belonging to two or more different angular profiles as they would cause inter-user interference, thereby leading to new angular profiles {tilde over (Φ)}i given by:
{tilde over (Φ)}i={(kj,lj)∉Φi:(kj,lj)∈Φi′∀i′≠i},i=0, . . . ,M′−1,where M′ denotes the number of different angular profiles after excluding beams causing inter-user interference, which coincides with the number of simultaneously addressable users;
allocating powers to each of the beams contained in said angular profiles {tilde over (Φ)}i, in such a way that the sum of powers of the beams in an angular profile does not exceed the power allocated to the corresponding user; and calculating excitation coefficients for each one of said large number of antennas by means of:AT[n,m,f]=1N1N2∑(k,l)∈Φ~PT[k,l]·(∑(k′,l′)∈Φ~U[k,l,k′,l′,f]ST[k′,l′,f])×exp(-j2πN1nk)exp(-j2πN2ml)where AT[n,m,f] is the excitation coefficient of antenna (n, m) in the frequency domain, U[k,l,k′,l′,f] are the elements of a precoding matrix in the frequency domain, PT[k,l] is the transmit power allocated to the beam in the direction (k, l), ST[k,l,f] is the data corresponding to the direction (k, l) in the frequency domain, and {tilde over (Φ)} denotes the superposition of the angular profiles of all the users with a number of elements, or beams, given byM≥M′:Φ~≡⋃i=0M′-1Φ~i.