Spectrum sharing is an important technique to improve the spectrum efficiency of dynamic ad hoc networks. However, since the BUSY/IDLE state of the ad hoc traffic could vary quickly, it is difficult to reduce the traffic collisions between coexisting networks. This chapter investigates a spectrum-sharing scenario, where a cooperative relay network intends to access the spectral band of a dynamic ad hoc network without making too many collisions with the ad hoc traffic. Adopting a binary continuous-time Markov chain (CTMC) traffic model to characterise and predict the ad hoc traffic, the spectrum access design of the cooperative relay network is formulated as a non-convex optimisation problem. We simplify this spectrum access design problem as a convex optimisation problem, which allows us to obtain a low-complexity optimal spectrum access strategy. Moreover, owing to the dynamic nature of the ad hoc traffic, the optimal spectrum access strategy needs to be attained within a short time and under practical limitations. To this end, we generalise the spectrum access design to an ergodic setting and propose an online spectrum access strategy. This online spectrum access strategy is able to achieve the optimal ergodic performance with negligible control delay, moderate signalling overhead and little computational capability requirement for the user equipment.
Dynamic spectrum sharing between cooperative relay and ad hoc networks: towards real-time optimal control, Page 1 of 2
< Previous page Next page > /docserver/preview/fulltext/books/te/pbte059e/PBTE059E_ch15-1.gif /docserver/preview/fulltext/books/te/pbte059e/PBTE059E_ch15-2.gif