UC San Diego

Wireless ad hoc networks have gained popularity in recent years. A networks' ability to self-organize without a centralized base station and the ease of deployment in a rough terrain make ad hoc networks desirable. Furthermore, for small mobile devices, it is not practical to mount multiple antennas to exploit the benefits of spatial diversity for combating multipath fading. With ad hoc networks, it is possible to take advantage of spatial diversity as nodes may cooperatively act together as if they were multiple antennas. For this dissertation, the system model is first presented to obtain the information- theoretic upper bound for peer-to-peer ad hoc network with hop-by-hop routing. This bound is not physically achievable, but is useful as a yardstick against any node cooperation scheme studied. Using the same system model, lower bound with node cooperation is studied in depth for two cases: simple time-share bound and frequency reuse bound. It is found that having more than two nodes cooperatively transmit together as a relay node does not provide significant improvement over two nodes cooperatively transmitting. The results also indicate that a more elegant scheme must be used for the frequency reuse case before the lower bound with node cooperation could match the information-theoretic upper bound. The simple scheme used here does not employ interference avoidance strategies beyond selecting nodes which are simultaneously sending must meet certain interference criteria. The results of using this scheme indicate that at a high signal to noise ratio, the bound with cooperation approaches peer-to-peer bound with no cooperation