On a Decentralized Deterministic Transmission Power Level Selection Algorithm in Large Aloha Networks under Saturation

In this paper, we will analyze a decentralized transmission power level selection algorithm in conjunction with Binary Exponential Back-off retransmission scheme, in large slotted-Aloha networks under saturation condition. In this algorithm, the transmission power is steadily increased by an amount called power step, until the packet is successfully transmitted. We will use SIR-based capture model to study the performance of this algorithm with small and large power steps. Our analysis proves that, although small power steps may seem favorable for power saving purposes, these power steps face throughput collapse under high load conditions (large number of nodes), while properly chosen large power steps guarantee higher throughputs and require approximately the same average powers.