|
The capacity of a wireless point-to-point, for example from a base station to a mobile phone/terminal, communication link to provide a particular quality of service is limited by the time-varying nature of the channel. Such variation is termed fading and is a consequence of the transmitted energy from the base station taking multiple paths to reach the mobile phone, as a result of, for example, reflections off building and other objects. At the mobile phone, when the energy from these different paths is combined, it will sometimes be constructive and lead to improved transmission, but, problematically, it will also be destructive, and result in a so-called deep fade, which can cause a complete break in transmission. These effects are time-varying due to the motion of the mobile. To overcome fading, some type of diversity must be introduced into transmission. In the standards for emerging broadband wireless systems such as WiFi (802.11) and WiMAX (802.16), it is proposed that multiple antennas could be used at both the base station and mobile device. Such a system is termed multiple-input multiple-output (MIMO) and potentially provides multiple different links between the transmitter and receiver, so that should one link be in fade, another will compensate, thereby ensuring more robust transmission. This performance gain is critically dependent upon appropriate positioning of the antennas to ensure that the multiple links have different fading properties, which might not always be possible due to the physical size constraints. To overcome this problem in this research, a multiple antenna transmitter or receiver will be formed by sharing the resources of a number of simple spatially separated single antenna nodes, an operation which is termed collaborative processing. Moreover, the single point-to-point link will be replaced by multiple physically shorter links, so that the required transmission energy is generally reduced, and a multihop network results. Transmission over each of these shorter links will be performed in a closed-loop manner, whereby information obtained at the receiver will be fed back to the transmitter, so that transmission can be performed over a link with the most favourable channel characteristics. Calculation of such favourable channels will be made possible by world-leading research in mathematical linear algebra, in particular polynomial matrix decompositions, unique to the research team in the proposal. Strategies for allocating the resources amongst the nodes, such as power, will also be designed, as will methods to mitigate interfence due to multiple users operating within the environment. In summary, therefore, the research project will focus on providing new algorithmic solutions for the application of the novel polynomial matrix decompositions with application to transmission over MIMO frequency selective wireless channels through academic partnership between Cardiff and Loughborough Universities, and industrial support from QinetiQ.
|