Marc Torrellas
Research Subjects
Marc Torrellas has a total of 2 co-authors. Here are the top co-authors by number of shared publications:
- Adrian Agustin
Universitat Politècnica de Catalunya (UPC) Barcelona
4 shared publications - Josep Vidal
Department of Signal Theory and Communications, Universitat Politcnica de CatalunyaBarcelonaTech (UPC), Barcelona, Spain
4 shared publications
Latest Publications
- Open Access
DoF-Delay Trade-Off for the $K$-user MIMO Interference Channel With Delayed CSIT
The degrees of freedom (DoF) of the $K$-user multiple-input multiple-output (MIMO) interference channel are studied when perfect, but delayed channel state information is available at the transmitter side (delayed CSIT). Recent works have proposed schemes that achieve increasing DoF values, but at the cost of long communication delays. This work proposes three linear precoding strategies, formulated in such a way that the achievable DoF can be derived as a function of the transmission delay, thus elucidating its achievable DoF-delay trade-off. All strategies are based on the concept of interference alignment, and built upon three main ingredients: delayed CSIT precoding, user scheduling, and redundancy transmission. In this respect, the interference alignment is realized by exploiting delayed CSIT in order to align the interference at the non-intended receivers along the space-time domain. Finally, the latter part of this work settles that all the proposed strategies work also for constant channels, except for SISO. In such a case, the schemes can be made feasible by resorting to asymmetric complex signaling concepts. This conclusion removes the time-varying channels assumption unnecessarily common along all the literature on delayed CSIT.
Open AccessDoF-Delay Trade-Off for the $K$-user MIMO Interference Channel With Delayed CSIT
The degrees of freedom (DoF) of the $K$-user multiple-input multiple-output (MIMO) interference channel are studied when perfect, but delayed channel state information is available at the transmitter side (delayed CSIT). Recent works have proposed schemes that achieve increasing DoF values, but at the cost of long communication delays. This work proposes three linear precoding strategies, formulated in such a way that the achievable DoF can be derived as a function of the transmission delay, thus elucidating its achievable DoF-delay trade-off. All strategies are based on the concept of interference alignment, and built upon three main ingredients: delayed CSIT precoding, user scheduling, and redundancy transmission. In this respect, the interference alignment is realized by exploiting delayed CSIT in order to align the interference at the non-intended receivers along the space-time domain. Finally, the latter part of this work settles that all the proposed strategies work also for constant channels, except for SISO. In such a case, the schemes can be made feasible by resorting to asymmetric complex signaling concepts. This conclusion removes the time-varying channels assumption unnecessarily common along all the literature on delayed CSIT.
- Open Access
Retrospective Interference Alignment for the MIMO Interference Broadcast Channel
The degrees of freedom (DoF) of the multiple-input multiple-output (MIMO) Interference Broadcast Channel (IBC) with 2 cells and 2 users per cell are investigated when only delayed channel state information is available at the transmitter side (delayed CSIT). Retrospective Interference Alignment has shown the benefits in terms of DoF of exploiting delayed CSIT for interference, broadcast and also for the IBC. However, previous works studying the IBC with delayed CSIT do not exploit the fact that the users of each cell are served by a common transmitter. This work presents a four-phase precoding strategy taking this into consideration. Assuming that transmitters and receivers are equipped with $M,N$ antennas, respectively, new DoF inner bounds are proposed, outperforming the existing ones for $\rho = \frac{M}{N} > 2.6413$.Open AccessRetrospective Interference Alignment for the MIMO Interference Broadcast Channel
The degrees of freedom (DoF) of the multiple-input multiple-output (MIMO) Interference Broadcast Channel (IBC) with 2 cells and 2 users per cell are investigated when only delayed channel state information is available at the transmitter side (delayed CSIT). Retrospective Interference Alignment has shown the benefits in terms of DoF of exploiting delayed CSIT for interference, broadcast and also for the IBC. However, previous works studying the IBC with delayed CSIT do not exploit the fact that the users of each cell are served by a common transmitter. This work presents a four-phase precoding strategy taking this into consideration. Assuming that transmitters and receivers are equipped with $M,N$ antennas, respectively, new DoF inner bounds are proposed, outperforming the existing ones for $\rho = \frac{M}{N} > 2.6413$. - Open Access
Retrospective Interference Alignment for the 3-user MIMO Interference Channel with delayed CSIT
The degrees of freedom (DoF) of the 3-user multiple input multiple output interference channel (3-user MIMO IC) are investigated where there is delayed channel state information at the transmitters (dCSIT). We generalize the ideas of Maleki et al. about {\it Retrospective Interference Alignment (RIA)} to be applied to the MIMO IC, where transmitters and receivers are equipped with $(M,N)$ antennas, respectively. We propose a two-phase transmission scheme where the number of slots per phase and number of transmitted symbols are optimized by solving a maximization problem. Finally, we review the existing achievable DoF results in the literature as a function of the ratio between transmitting and receiving antennas $\rho=M/N$. The proposed scheme improves all other strategies when $\rho \in \left(\frac{1}{2}, \frac{31}{32} \right]$.Open AccessRetrospective Interference Alignment for the 3-user MIMO Interference Channel with delayed CSIT
The degrees of freedom (DoF) of the 3-user multiple input multiple output interference channel (3-user MIMO IC) are investigated where there is delayed channel state information at the transmitters (dCSIT). We generalize the ideas of Maleki et al. about {\it Retrospective Interference Alignment (RIA)} to be applied to the MIMO IC, where transmitters and receivers are equipped with $(M,N)$ antennas, respectively. We propose a two-phase transmission scheme where the number of slots per phase and number of transmitted symbols are optimized by solving a maximization problem. Finally, we review the existing achievable DoF results in the literature as a function of the ratio between transmitting and receiving antennas $\rho=M/N$. The proposed scheme improves all other strategies when $\rho \in \left(\frac{1}{2}, \frac{31}{32} \right]$. - Open Access
On the Degrees of freedom of the K-user MISO Interference Channel with imperfect delayed CSIT
This work investigates the degrees of freedom (DoF) of the K-user multiple-input single-output (MISO) interference channel (IC) with imperfect delayed channel state information at the transmitters (dCSIT). For this setting, new DoF inner bonds are provided, and benchmarked with cooperation-based outer bounds. The achievability result is based on a precoding scheme that aligns the interfering received signals through time, exploiting the concept of Retrospective Interference Alignment (RIA). The proposed approach outperforms all previous known schemes. Furthermore, we study the proposed scheme under channel estimation errors (CEE) on the reported dCSIT, and derive a closed-form expression for the achievable DoF with imperfect dCSIT.
Open AccessOn the Degrees of freedom of the K-user MISO Interference Channel with imperfect delayed CSIT
This work investigates the degrees of freedom (DoF) of the K-user multiple-input single-output (MISO) interference channel (IC) with imperfect delayed channel state information at the transmitters (dCSIT). For this setting, new DoF inner bonds are provided, and benchmarked with cooperation-based outer bounds. The achievability result is based on a precoding scheme that aligns the interfering received signals through time, exploiting the concept of Retrospective Interference Alignment (RIA). The proposed approach outperforms all previous known schemes. Furthermore, we study the proposed scheme under channel estimation errors (CEE) on the reported dCSIT, and derive a closed-form expression for the achievable DoF with imperfect dCSIT.