In a Nutshell, Misc, Projects

Interference alignment in a nutshell

Jorge F. Schmidt explaining interference alignment

There are different strategies on how to handle and manage interference in wireless networks. The choice of strategy mainly depends on the signal-to-interference ratio (SIR) at the receiver. If interference is very weak, it should be ignored and treated as noise. If interference is stronger—about as strong as the desired signal—we want to avoid it and allow only one node to transmit at a given time or frequency in a certain spatial region (orthogonalization). Both strategies are commonly used in wireless technologies in practice. If interference is even stronger than the signal of interest, it is favorable to allow interference and decode it at the receiver rather than to avoid it. A major limitation of all three strategies is that their per-node throughput decreases with an increasing number of transmitting nodes.

This fundamental challenge is tackled by interference alignment, which is a “revolutionary wireless transmission strategy” (El Ayach) that gives a fresh perspective on capacity limits and has the potential to significantly improve the spectral efficiency of wireless communications. The basic idea is that transmissions of multiple senders are coordinated in a way that each receiver obtains the intended signal during half of the time (“everyone gets half the cake” (Cadambe)), and the accumulated interference (the signals from all other senders) occupies the other half. In general terms, each sender encodes its signals over multiple dimensions of the signal space. To be able to recover its signal, two conditions must hold: its own signal space must not be inside the signal space spanned by all the other senders and the dimension of the signal space spanned by all other senders must be strictly smaller than the number of measurements. When these conditions hold, all interference signals occupy a low-dimensional signal space, and the desired signal can be recovered at the receiver by projecting the received samples into the null space of the interference. The great advantage is that—in contrast to traditional orthogonal multiple access—the capacity is no longer the limiting factor in a multiple node scenario, but the network throughput increases linearly with the number of nodes.

Interference channel

Interference channel

Different flavors of interference alignment have been investigated since its invention seven years ago. They can be classified according to their level of channel state information (CSI) required at the transmitter. There are schemes requiring global CSI, retrospective schemes that can cope with outdated CSI, and completely blind schemes. Researchers have investigated the feasibility conditions of interference alignment on its different variations and developed computationally efficient schemes, such as iterative alignment and distributed alignment. The design of practically feasible alignment techniques is still at an early state. Open issues include the following: How to handle the feedback requirement for global CSI over time-varying channels? How to apply interference alignment in large networks?

Literature

  • S. Bazzi, G. Dietl, W. Utschick, “Large system analysis of interference alignment achievable rates for the MIMO interference channel,” IEEE Trans. Signal Process., 63(6):1490–1499, 2015.
  • V. R. Cadambe, S. A. Jafar, “Interference alignment and the degrees of freedom of the K-user interference channel,” IEEE Trans. Inf. Theory, 54(8):3425–3441, 2008.
  • O. El Ayach, S. Peters, R. Heath, “The practical challenges of interference alignment,” IEEE Wireless Commun. Mag., 20(1):35–42, 2013.
  • S. A. Jafar, “Blind interference alignment,” IEEE J. Sel. Topics Signal Process., 6(3):216–227, 2012.
  • M. Maddah-Ali, A. Motahari, A. Khandani, “Communication over MIMO X channels: Interference alignment, decomposition, and performance analysis,” IEEE Trans. Inf. Theory, 54(8):3457–3470, 2008.
  • H. Maleki, S. A. Jafar, S. Shamai, “Retrospective interference alignment over interference networks,” IEEE J. Sel. Topics Signal Process., 6(3):228–240, 2012.
  • K. Miller, A. Sanne, K. Srinivasan, S. Vishwanath, “Enabling real-time interference alignment: Promises and challenges,” in Proc. ACM MobiHoc, pp. 55–64, 2012.
  • Y. Wu, S. Shamai Shitz, S. Verdu, “Information dimension and the degrees of freedom of the interference channel,” IEEE Trans. Inf. Theory, 61(1):256–279, 2015.
  • G. Zheng, I. Krikidis, C. Masouros, S. Timotheou, D.-A. Toumpakaris, Z. Ding, “Rethinking the role of interference in wireless networks,” IEEE Commun. Mag., 52(11):152–158, 2014.

The text of this blog entry was written together with Jorge F. Schmidt.