Adaptive Two-step Binary Exponential Backoff Strategy for Random Access

In the process of random access, the backoff algorithm is a very common congestion control method. A reasonable backoff algorithm can effectively alleviate system congestion. However, due to changes in the network state, the backoff algorithm cannot quickly select the most suitable backoff window. This may lead to oscillation of backoff window length. In this paper, an adaptive backoff algorithm is proposed to dynamically adjust the backoff window based on the current number of loads. The backoff window length is adjusted after the transmission of data packet. When the number of loads is small, the node will have a high probability of resetting the window length to the initial value. When the number of loads is large, the node will have a greater probability of maintaining the current backoff window length. Consequently, the proposed algorithm could avoid the occurrence of doubling the backoff window length multiple times. Simulations are performed to verify the superiority of the proposed algorithm.