For many wireless communication links, such as those employing turbo codes or sequentially-decoded convolutional codes, the power consumption of the decoder at the receiver depends on the received signal power and, hence, on the transmitted signal power. By transmitting a signal at a power higher than the minimum required for successful reception, the transmitter can "assist" the receiver by reducing the receiver decoding effort and thereby reducing the overall power consumption on the link. In this paper, we consider sender-receiver power tradeoffs in data-gathering trees. We formulate an optimization problem to optimally assign power to the nodes in the tree for maximizing the lifetime of the data-gathering tree, which is equivalent to the time until network partition due to battery outage. We propose a Binary Search Algorithm for optimal power assignment among nodes that maximizes the tree lifetime. Our Binary Search Algorithm can be easily extended to handle practical considerations such as node mobility and peak power constraints at nodes. Using turbo codes as an example of a channel coding technique, we demonstrate significant improvements in network lifetimes as a result of sender-receiver power tradeoffs. These improvements are observed under a wide variety of network conditions and are more pronounced in densely deployed networks and networks with asymmetric power costs.
展开▼
