Abstract: A time variant power allocation (Time Variant Power Allocation, TVPA) algorithm is proposed for the asymmetric wireless cooperative communication network, which is affected by reflecting and scattering. According to the time variant and asymmetric channel conditions between different nodes in the wireless cooperative network, the algorithm distributes the transmitted signal power between the source node and relay node during the signal transmission. The non-convex optimization problem of minimizing the system error probability is converted into the convex optimization problem of maximizing the system capacity by using the channel coding theory. Compared with the fixed power allocation (FXPA) algorithm and average power allocation (AVPA) algorithm, the proposed algorithm can utilize the channel variety efficiently, and reduce the system error probability by allocating the transmitted power of each node adaptively. The decode-and-forward cooperative scheme and the quasistatic Rayleigh fading channel are considered in this paper. Simulation results in various network models show that, compared to FXPA algorithm, TVPA algorithm can achieve about 5.5dB bit error rate gains. This superiority increases as the network quality increases.