Subspace Alignment Chains and the Degrees of Freedom of the Three-User MIMO Interference Channel

• Published in: IEEE transactions on information theory Vol. 60; no. 5; pp. 2432 - 2479
• Main Authors: Chenwei Wang, Tiangao Gou, Jafar, Syed Ali
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.05.2014; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Alignment; Antennas; Channels; Degrees of freedom; Electrical engineering; Interference; Interference channels; MIMO; Receivers; Receiving antennas; Redundancy; Subspaces; Time-frequency analysis; Transmitters; interference alignment; MIMO interference channel; degrees of freedom; subspace alignment chains
• ISSN: 0018-9448; 1557-9654
• DOI: 10.1109/TIT.2014.2308988

We show that the three-user M T × M R Multiple-Input Multiple-Output (MIMO) interference channel where each transmitter is equipped with M T antennas and each receiver is equipped with M R antennas has (M) d(M, N) = Δ min (M/2-1/κ, N/2+1/κ) degrees of freedom (DoF) normalized by time, frequency, and space dimensions, where = Δ min(M T , M R ), N = Δ max(M T , M R ), κ = Δ ⌈M/N-M⌉. While the DoF outer bound of d(M, N) is established for every M T , M R value, the achievability of d(M, N) DoF is established in general subject to a normalization with respect to spatial extensions, i.e., the scaling of the number of antennas at all nodes. In particular, we show that qd(M, N) DoF are achievable for the three-user qMT × qMR MIMO interference channel, for some positive integer q, which may be seen as a spatial extension factor. q is the scaling factor needed to make the value qd(M, N) an integer. Given spatial extensions, the achievability relies only on linear beamforming based interference alignment schemes and requires neither channel extensions nor channel variations in time or frequency. In the absence of spatial extensions, it is shown through examples how essentially the same interference alignment scheme may be applied over time-extensions over either constant or time-varying channels. The central new insight to emerge from this paper is the notion of subspace alignment chains as the DoF bottlenecks. The subspace alignment chains are instrumental both in identifying the extra dimensions to be provided by a genie to a receiver for the DoF outer bound, as well as in the construction of the optimal interference alignment schemes. The DoF value d(M, N) is a piecewise linear function of M, N, with either M or N being the bottleneck within each linear segment, whereas the other value contains some redundancy, i.e., it can be reduced without reducing the DoF. The corner points of these piecewise linear segments correspond to two sets, A={1/2, 2/3, 3/4, ...}and B={1/3, 3/5, 5/7, ...}. The set A contains all those values of M/N and only those values of M/N for which there is redundancy in both M and N, i.e., either can be reduced without reducing the DoF. The set B contains all those values of M/N and only those values of M/N for which there is no redundancy in either M or N, i.e., neither can be reduced without reducing the DoF. Because A and B represent settings with maximum and minimum redundancy, essentially they are the basis for the DoF outer bounds and inner bounds, respectively. Our results settle the question of feasibility of linear interference alignment, introduced previously by Cenk et al., for the three-user M T × M R MIMO interference channel, completely for all values of M T , M R . In particular, we show that the linear interference alignment problem (M T × M R , d) 3 (as defined in previous paper by Cenk et al.) is feasible if and only if d ≤ d(M, N). With the exception of the values M/N ∈ B, and only with that exception, we show that for every M/N value there are proper systems (as defined by Cenk et al.) that are not feasible. Evidently the redundancy contained in all other values of M/N manifests itself as superfluous variables that are not discounted in the definition of proper systems, thus creating a discrepancy between proper and feasible systems. Our results show that M/N ∈ A are the only values for which there is no DoF benefit of joint processing among co-located antennas at the transmitters or receivers. This may also be seen as a consequence of the maximum redundancy in the M/N ∈ A settings.