Information and Signal Processing for Intelligent Communications (ISPIC)


We are a group of researchers in advanced communications and information systems. We conduct theoretical and experimental research aligned with the following areas of technology: Artificial intelligence for wireless communications, signal processing for large/distributed/multi-user antenna systems, signal processing and coding for future wireless networks (5G, 6G, and beyond), (deep) space communications  and coding for computing at the network edge.

Our theoretical research work is aligned with the following fields of fundamental research (information and communication theory, advanced signal processing, and artificial intelligence) and is complemented with the experimental activity conducted in the AI Wireless Lab.

The Lab capabilities include: over-the-air simulation of wireless communication systems, demonstration of AI/ML aided communication algorithms, proofs-of-concept of 5G/6G physical layer blocks and GPU-aided computations (for machine learning simulations).

Research lines

While traditional, model-driven engineering approaches to communication systems design reach their limits, there is an increasing need for customizing data-driven algorithms to the physical layer. The benefits of AI/ML algorithms are an increased robustness, versatility and adaptivity to model changes, and the ability to find near-optimal solutions where analytical methods become intractable.

We aim to contribute to the evolution from state-of-the-art 5G systems towards a fully AI-native air interface as envisioned for 6G. Several research projects are being carried out at our lab, which cover the following topics:

  • End-to-end learning of physical-layer functions, error-correcting codes, and coded modulation
  • Data-driven beamforming and user clustering algorithms
  • Massive random-access systems for machine-type communications based on reinforcement learning
  • Routing algorithms for vehicular communications based on reinforcement learning
  • Emergent communications and automated MAC protocol learning

Traditionally, wireless communication systems have benefited from exploiting the spatial dimension to increase the system throughput by introducing multiple antennas at transceivers. Although this concept has been deeply investigated in the past, there are arising new concepts that deal with massive connectivity scenarios such as Extra Large Antenna Arrays (ELAA) and new access architectures such as cell-free Massive MIMO, all of which have a tremendous potential to influence next generation wireless systems. Having large antenna systems requires to work with large data sets where the sample size has the same order of magnitude as the number of variables. This is an important challenge that can be addressed by means of multivariate statistics theory, for example using the large-dimensional random matrix theory (RMT) framework applied to large antenna systems.

The following is a list of current research topic conducted in our group:

  • Signal processing and channel modeling for Extremely Large Antenna Arrays (ELAA)
  • Synchronization and channel estimation for distributed and cell-free MIMO systems
  • Spectrum sensing in large dimensional settings
  • Massive connectivity

The variety of frequency bands and propagation environments that are to be employed in 6G systems impose the need for advanced waveforms and link adaptation strategies that can be dynamically adapted according to the user context (position, velocity, frequencies, services, etc.). One of the key features of 6G is expected to be the integration of communication and radio sensing capabilities in the same spectrum. Recent advances in mmWave, massive MIMO and machine learning are key to facilitate this integration, although many technical challenges remain, from waveform design to MAC optimization.

Our lab aims to contribute to this trend by addressing the following research topics:

  • Advanced signal processing techniques for communications
  • New waveforms, modulations, and coding algorithms
  • Optimization of coding and decoding schemes for next generation wireless systems
  • Joint sensing and communications
  • Coding for secure communications, secret key agreement and quantum key distribution

Future wireless networks like 6G are not only expected to provide connectivity to tens of billions of edge devices, but also to contribute to equipping them with computing capabilities and intelligence. Relevant scenarios are, for example, Industry 4.0 or connected autonomy. How to perform joint communication, computation and learning efficiently, considering resource limitations in terms of energy consumption, bandwidth, memory and computation resources, etc. as well as application specific requirements such as latency or privacy and security constraints, are the main objectives of this line of research.

Our approach leverages on information and coding theory. The current research topics at our lab include:

  • Over-the-air computing
  • Polynomial coding for private and secure distributed coded computing
  • Physical-layer network coding techniques
  • Coded caching