Carrier synchronization is critical to any implementation of distributed transmit beamforming. For a practical network, the transmit nodes are not tethered and each has its own local oscillator (LO). Due to the low cost of these off-the-shelf LOs, there is substantial frequency offsets which hinder carrier synchronization between cooperating transmitters. A robust algorithm to pre-synchronize the carriers of the transmitters is the so called two node consensus based synchronization. Here, a designated master node transmits a monotone, , to a node , where denotes the phase. Node synthesizes the received tone, which includes zero mean additive noise following the setup illustrated in Figure 1.
Figure 1: Carrier synchronization at node .
The algorithm implemented at node follows the following equations:
Here, is the instantaneous frequency from the LO of node , is a stability regularization term with being a descent step size. The above system is locally stable and the algorithm converges asymptotically leading to the paired transmitter coherence. In practice, carrier pre-synchronization is achieved before distributed beamforming to a far end receiver starts.