This paper reviews our recent research on the optimal power control at the physical layer, and scheduling at the medium access control (MAC) layer under energy-constraints. In our approach, first, by use of Nash game theory, the physical and link layer designs are jointly minimized to provide an optimal power control solution for multi-carrier (MC) direct sequence (DS) CDMA to maximize the utility function defined as the number of data bits transmitted correctly per unit of energy. That is, the network throughput is maximized at the cost of the minimal energy consumed. Second, the pair-wise optimization of the hardware and MAC layer designs is considered to present an optimal scheduling protocol to minimize the energy utilization for data fusion in wireless sensor network, which includes both the transmission energy and the circuit energy consumption. The integration via cross-layer protocol optimizes the communication system's use of energy while making the system more readily adaptive to changes in the environment and user tendencies. This is very crucial for the emerging IT technologies, since it is well known that the energy resource is getting more and more expensive nowadays.