Interference alignment with doubly layered signaling for constant SISO interference channels

It has been conjectured by Høst-Madsen and Nosratinia that the K-user single-input single-output (SISO) complex Gaussian interference channels with constant channel coefficients have merely one degree of freedom (DoF) regardless of the number of users, i.e., K. Then, Cadambe and Jafar introduced the idea of interference alignment (IA) being able to achieve K/2 DoF in time-varying SISO interference channels. Moreover, their joint work with Wang settled the Høst-Madsen-Nosratinia conjecture in negative by using the idea of asymmetric complex signaling to achieve 1.2 DoF for K-user constant SISO interference channels. In this paper, a linear IA scheme for K-user constant SISO interference channels is proposed which could enable us to achieve K/4 DoF for almost all channel coefficients. This means that whenever K ≥ 5, the proposed scheme could achieve at least 1.25 DoF. The main idea of the proposed method relies on the linear IA using symbol extension by Cadambe-Jafar which is not effective for constant channels. However, we show that along with signal rotation across every two consecutive time slots to artificially build a random time-varying channel out of a constant channel, the proposed method can be directly applied to constant channels to achieve K/4 DoF.

[1]  Aria Nosratinia,et al.  The multiplexing gain of wireless networks , 2005, Proceedings. International Symposium on Information Theory, 2005. ISIT 2005..

[2]  Syed Ali Jafar,et al.  Interference Alignment With Asymmetric Complex Signaling—Settling the Høst-Madsen–Nosratinia Conjecture , 2009, IEEE Transactions on Information Theory.

[3]  Syed Ali Jafar,et al.  Interference Alignment and Degrees of Freedom of the $K$-User Interference Channel , 2008, IEEE Transactions on Information Theory.

[4]  Erik Ordentlich,et al.  The Degrees-of-Freedom of the $K$-User Gaussian Interference Channel Is Discontinuous at Rational Channel Coefficients , 2009, IEEE Transactions on Information Theory.

[5]  Gene H. Golub,et al.  Matrix computations (3rd ed.) , 1996 .

[6]  Shlomo Shamai,et al.  Degrees of Freedom Region of the MIMO $X$ Channel , 2008, IEEE Transactions on Information Theory.

[7]  Robert H. Halstead,et al.  Matrix Computations , 2011, Encyclopedia of Parallel Computing.

[8]  Amir K. Khandani,et al.  Interference alignment for the K user MIMO interference channel , 2009, 2010 IEEE International Symposium on Information Theory.