Our efforts in the area of Medium Access Control (MAC) and scheduling are motivated by challenges introduced by the rapid evolution of wireless technologies and by the ongoing introduction of new applications. Our work includes both theoretical contributions aimed at furthering our understanding of the performance limits of wireless networks as well as practical aspects, in which we study and optimize the performance of emerging technologies.
Our ongoing efforts focus on:
- A cellular measurement campaign and the design of predictive scheduling algorithms for mobile users in cellular networks.
- Design and experimental evaluation of efficient multicast schemes for large crowds.
- Design scheduling algorithms for Cooperative Multipoint (CoMP) systems.
- Design of resource allocation algorithms for full duplex transceivers.
An important aspect of our research is the design of algorithms and architectures that take into account cross layer considerations. In particular, an important cross layer problem is the design of MAC protocols for emerging physical layer technologies. For example, we considered the implications of the multipacket reception capability which has the potential to improve the performance of the MAC layer. We showed that in networks with such a capability, the negative implications of reusing the legacy protocols include not only unfairness but also a significant throughput reduction. Moreover, in collaboration with Clariton Networks we analyzed a novel architecture and MAC modifications for a system that provides in-building cellular and Wi-Fi communication by utilizing the Community Antenna Television (CATV) infrastructure. Recently, similar systems became popular and are now known as a Distributed Antenna Systems.
Finally, our early work in the area of Bluetooth networks included the analysis of bandwidth allocation algorithms and of scheduling algorithms. For example, we studied the performance of the Bluetooth MAC protocol and obtained analytical results regarding the performance of intra-piconet scheduling algorithms by applying techniques from queueing analysis of polling systems.