Wireless networks

  • Device-to-device communications.We explore the feasibility of device-to-device communications when users are equipped with widely popular terminals as smartphones or tablets. We investigate different software solutions, such as WiFi Direct, and aim at developing new features that improve the capabilities of the technology and the services that the users can enjoy.
  • Cellular networks.We mainly address the problem of radio resource allocation in next-generation cellular networks. Such networks are foreseen to be heterogeneous as they make use of various wireless technologies and provide very different radio coverage (from macrocells to picocells). Through analytical and simulation techniques, we investigate the performance of such complex systems and envision optimal solutions for the support of mobile users and a vast range of services.
  • Mobile networks: a critical issue in mobile networks pertains to achieving a desired distribution of the information within an area, i.e., regardless of how the information is distributed at the outset, the system should be able to identify where the information should be stored in the network area. Also, regardless of the initial information distribution and of the density of nodes, information should never be allowed to die out. We address these issues by an innovative approach which lets the information move across the network nodes, and the network nodes replicate or remove the stored information as needed. Another line of research investigates the performance of multicasting in static as well as mobile multihop networks, considering also network coding techniques .
  • Delay Tolerant Networks (also referred as Opportunistic Networks): DTN are an emerging packet switching paradigm, in which no fixed infrastructure is required to communicate between any nodes, even if the network is at any time partitioned. Data routing is based essentially on exploiting: i) the cooperation among the nodes, which can transfer data through intermediate relays in multi-hop fashion, ii) their physical mobility, which enables them to carry physically data from one location to another and meet the destination or some relay. Research activities are focused on many aspects: i) theory to understand the performance scaling laws, (ii) algorithm design to compute the optimal routing and (iii) implementation exploiting Android portable devices.
  • Vehicular networks: information dissemination and caching are aspects of paramount importance in mobile multihop networks and, in particular, in vehicular networks. Our focus is on the support of publish/subscribe-based services in a distributed environment such as a multihop network. The issues under study include efficient query propagation, limiting information storage within a target geographical area, and caching strategy in an environment where a cache-all-you-see approach is clearly unfeasible but where the availability of sought-after information from nearby nodes is the key to success. We also address security issues, so as to ensure that the actual vehicles positions can be retrieved.
  • Mesh networks: we address issues pertaining to persisting connections and reconfiguration problems around broken paths by selecting alternate routes from node to node, and taking hop-by-hop routing decisions, until the destination is reached. Specifically, we aim at solving the problem through:
    • Cross-layer routing metrics: routing algorithms carrying the knowledge of transmission rates and link qualities of all mesh nodes sharing the same channel.
    • Frequency assignment algorithms: this problem, although extensively studied in the literature, is given a new spin in real implementations due to interference coming from privately-owned Access Points. We seek to build adaptability and re-computation in the design of frequency assignment algorithms for real mesh networks.
  • Wireless sensor networks for event detection and measurement collection: we study the ability of single- and multi-hop sensor networks to convey measurements as well as alarm messages to a central controller. Our study is based on analytical techniques, which include the application of random matrix theory. Through our models, we consider the reconstruction of the phenomenon under observation at the central controller from the set of irregular samples collected through the sensors, and we study the quality of the reconstructed signal.
  • Fundamental scaling laws in static wireless networks with inhomogeneous node density: wireless ad-hoc networks have been traditionally modeled as a set of nodes placed according to a Poisson process over a finite bi-dimensional domain and communicating among them (possibly in a multi-hop fashion ) over point-to-point wireless links subject to mutual interference. The goal of our research activity is to extend the capacity scaling analysis to networks exhibiting a much higher variability in the node spatial distribution than the one resulting from a homogeneous Poisson point process. Indeed, almost all large-scale structures created by human or natural processes over geographical distances (such as urban or sub-urban settlements) are characterized by significant degrees of clustering, due to spontaneous grouping of the nodes around a few attraction points.


Research groups