Modeling And Mitigating Interference For Wireless Ad Hoc Networks

One of the major factors limiting capacity in wireless networks is radio interference. Dealing with radio interference will become increasingly important in wireless ad hoc networks, as scenarios in which several closely located wireless nodes sharing the same frequency bands for multi-hop communication. Interference can degrade the performances of wireless ad hoc networks, such as network throughput, energy consumption, network lifetime and other aspects.The above discussion motivates the considerable interest of the research community in understanding and analyzing the impact of interference on wireless ad hoc network performance.A major challenge in dealing with wireless interference is how to model such a complex physical phenomenon. In fact, the previous interference models are not accurate and manageable from an analytical/algorithmic point of view. In this part of the dissertation, we classify these loss frames into two categories: collision and interference. Moreover, interference is divided into synchronous and asynchronous interference. Synchronous interference is that one or several nodes start to transmit at the same physical slot and they cause the reference signal frame to be corrupted. In asynchronous interference, the transmission node can not sense that some nodes have been transmitting in the interference area.Another fundamental challenge in the design of "interference-aware" network protocols is how to mitigate or eliminate interference to approach to the optimal performance in wireless ad hoc networks. In this part of the dissertation, we present three cross-layers of interference-aware approachs for wireless ad hoc networks.1. Channel Interference and Access Delay (CIAD) routing meric. In ad hoc networks any concurrent transmission will add to the level of interference and collision experienced by other users. The spatial density of transmissions that could be supported subject to two constraints: the excessive interference of attempted receivers, the channel access collision of attempted transmitters. We combine two constraint factors and define Channel Interference and Access Delay routing metric. CIAD may balance interference, access delay and length of a route. CIAD can avoid accessing the areas with high interference and overfull channel competition. CIAD increases throughput, reduces the end-to-end delay and packets loss ratio.2. Interference-aware Physical Carrier Sensing Protocol. Physical carrier sensing threshold plays an important role in the channel access activities. As the channel access activities and packet receive rate of a node is affected by the accumulative interference, we present a new analytical approach that demonstrates how to tune the sense threshold to reduce the interference. Then we present a new heuristic algorithm to optimize the physical carrier sense threshold by measuring the state of the channel in IEEE 802.11 ad hoc networks. The optimum carrier sensing range should balance the spatial reuse and the impact of interference in order to optimize the aggregate throughput of nodes (users).3. Hybrid Interference Model-based Topology Control Algorithm (HIMTC). All previous approaches used for topology control have in common that they model wireless networks as purely geometric graphs, hence neglecting one of the most crucial aspects of wireless communication: capture effect. From a theoretical point of view, the graph-based models of topology control really don't bear any significance when it comes to actually scheduling messages in an SINR environment. Therefore, we integrate SINR-based model into graph-based model. Then we present a hybrid interference model-based topology control algorithm. The proposed algorithm is distributed and only requires the local information. It can lower the graph interference of networks and increases the network capacity.