Swing

Context

The tremendous growing needs of radio resources associated with the recent explosion of wireless-based services led the standardization industry to propose a large number of standards driven by different groups from the IEEE (802 family), ETSI (GSM), 3GPP (3G, 4G) or the Internet Society (IETF standards). Roughly speaking, the upper layers of the OSI model, from layer 3 (network) up to layer 7 (application) are nowadays mostly driven by the proposals of the IETF, while the IEEE802 family offers the major contributions at layers 1 and 2 (except for cellular networks) and represents the basis of numerous famous commercially available technologies such as WiFi (802.11), Wimax (802.16), Bluetooth or Zigbee (802.15.4),... So far, all standards are developped with a bottom-up approach: physical (PHY) and radio link (RL) layers are firstly defined (see the IEEE 802 family for instance) and provide specific capacity and multiplexing capabilities. Then, the networking (NET) layer is derived according to the IETF proposals or some other consortia such as theWiFi alliance or theWimax forum. Of course, it is worth mentioning that application needs are often accounted for partly for the design of the lower layers, but since these needs change rapidly, they appear to be obsolete when a standard is finalized. The way radio technologies have been developed up to now is far from benefiting a real wireless convergence of the system. Thus, the current development of the wireless industry could be slowed down by the lack of radio resources and adaptability of current wireless technologies. To cope with the future needs of wireless systems, the research commmunity will have to face three types of challenges : terminal flexibility, agile radio resource management and autonomous networking. standards

Scientific foundation

Flexible radio node design

Designing a radio node is definitely not an all-analogical process. Since software defined radio principles were established, some new features such as adaptability and auto-reconfiguration are becoming mandatory for the terminal to adapt to its environment and to the application in use. This implies that an important part of the radio coding/ decoding process is done in the digital world. Because a full software radio node is still an utopia, future architectures will have to cope with analogical and digital constraints and their co-design is a real challenge. New computation models are emerging, such as for instance, the concept of radio virtual machine or new hardware abstraction layers permitting to develop separately, radio protocols, strategy for resource sharing, operating systems and top-level applications.

Agile radio resource sharing

Radio resource sharing is very important in autonomous and spontaneous networks. This problem covers several research fields including signal processing and protocols. In various contexts from wireless sensor networks (WSNs) up to cellular wireless networks, the problem of assessing the radio resource remains a challenging issue. Mitigating interference for multi-system environments, optimizing energy and capacity for high data rate access networks or increasing the life-time of WSNs all strongly rely on the resource sharing strategy. The complexity of this problem originates from the inherent properties of the radio channel itself which is subject to highly variable propagation phenomena and interference. Because radio environments are dynamics, as well as the users and QoS needs, future systems will have to integrate self-adaptive, real-time and distributed algorithms. More recently, a tremendous interest for cooperative techniques appeared, that allow the nodes to do more than coexisting: they can cooperate. This is a very competitive issue especially for heterogeneous systems, when the nodes only have a partial view of their radio environment. This cooperation can be addressed at the signal level (virtual MIMO) or at the coding level (network coding), in a strong relationship with the data link layer to ensure robustness of end to end communications.

Autonomous wireless networking

The previously described mechanisms allow to manage efficiently the radio resource in the neighbourhood of a node by taking into account the different wireless interactions. Now, the objective is to route a data from a source to a destination. This well-known problem should be revisited in the context of wireless networks, and more particularly if we want to take benefit from agile radio, opportunistic radio links, non-symmetric neighbours and so on. Because of the large-scale dimension of the networks we consider, centralized approaches should be dismissed to the benefit of the development of distributed and localized protocols: based on local information and local interactions, the aim is to synthesize a global behaviour in terms of routing, data gathering, etc. The most important issues deal with activity scheduling, topology control and protocols adaptability to the evolution of the network topology. Because such features need to be human-free, they are often referred to as the self-* paradigm which will drive our research effort. Further, since network topologies are constantly evolving due to the mobility of the nodes and the variability of the radio links properties, fault-tolerant protocols are needed to guarantee robustness and self-stabilization.

People

Head

Faculty

Guests, exteriors and assistants

PhD students