cross_layer_optimization_in_wireless_multi_hop_networks

Cross-Layer Optimization in Wireless Multi-Hop Networks

Cross-layer design becomes a more and more important topic for wireless networks, since the layer based improvements are reaching the achievable limits. Techniques like OFDMA and MIMO increased the efficiency and reliability of the PHY. In combination with a channel aware scheduler at the MAC, the system performance can be enhanced even further. It can be foreseen that the future wireless network is going to be a heterogeneous wireless multi-hop network, where mobile subscribers with different service needs (e.g. video-telephony, streaming, etc.) and the internet of things (e.g. machine-to-machine, wireless sensor networks, etc.) coexist in one and the same network. Hence, cross-layer driven optimization for wireless multi-hop networks has to be extended to the NET layer in order to improve the system performance.

In this work, we are considering a cross-layer optimization on the PHY/MAC/NET layers, where the inherent properties of the wireless multi-hop network are utilized at PHY and also at the MAC and the NET. Such inherent properties are for instance the channel variations in time and frequency as well as the broadcast nature of the radio transmission. This PHY knowledge shall be used at the NET layer to improve the routing performance. As another example, the routing paths could be chosen such that physical layer network coding is enabled and the respective gains can be leveraged.

Investigations are done for unicast and multicast transmission and cooperative and non-cooperative communication schemes and their combinations on PHY/MAC/NET. During this work, a non-ideal environment is assumed, which means that imperfect channel knowledge, efficiency loss due to signaling and so on will be taken into account. Optimization goals considered in this project are efficiency, robustness and fairness of the system. Evaluation criteria are e.g. throughput, energy consumption, delay, etc.

Besides conventional optimization techniques (e.g. convex optimization), also game theory is used to obtain results for this project. The interest in game theory for wireless communication is due to the possibility to model autonomously acting nodes. If applicable to the respective scenario, stable points of a system with autonomous nodes can be obtained and in the case of cooperative nodes, the social optimum of such system can be achieved in a distributed manner. Further, game theory can be used to introduce incentive mechanisms in such a way that wireless nodes will behave in favor of the overall system performance.